Tue 23 June 2026
Daily Brief — Curated and contextualised by Best Practice AI
The stories that matter most
Selected and contextualised by the Best Practice AI team
Qualcomm Nears Deal for AI Chip Startup Modular
Qualcomm Inc. is in advanced talks to acquire Modular Inc. in a transaction valuing the artificial intelligence infrastructure software company at about $4 billion, according to people familiar with the matter.
Is China Closing the A.I. Gap Faster Than Expected?
Silicon Valley and corporate America are increasingly turning to cheaper, open-source artificial intelligence models built in China.
Bain tests software takeover targets by vibecoding AI replicas
Private equity groups swiftly recreate software products to gauge their competitive advantages
South Korean chipmakers are being paid such massive bonuses it’s becoming an inflation problem for the central bank
The Bank of Korea is worried that a wage-price spiral is increasing consumer demand and pushing up prices.
Gas Stations Accused of Using AI to Boost California Prices
A group of California consumers claimed in a lawsuit that gas station owners including Walmart Inc., Marathon Petroleum Corp., BP Plc and 7-Eleven Inc. are using artificial intelligence to illegally manipulate pump prices in the state that already has the highest rates in the US.
AI Employment Impact: AI boom's US employment, wage impact muted so far, ECB study finds, ETHRWorld
AI Employment Impact: Firms have been investing heavily in AI in recent years, raising fears that humans will be replaced at increasing rates, curbing overall employment and widening inequality along the way.
No Claude Fable 5? No problem: Sakana achieves frontier performance with new Fugu multi-model, auto synthesis system
Last night, the increasingly enterprise-focused AI startup Sakana launched Fugu, a multi-agent orchestration system that delivers frontier-level AI performance through a single, OpenAI-compatible API. Designed for developers, enterprises, and nations seeking resilience against vendor lock-in and geopolitical export controls, Fugu (Japanese for "pufferfish"), bypasses the traditional monolithic model structure by dynamically routing queries to a swappable pool of specialized AI agents. Sakana CEO and co-founder David Ha, formerly of Google Brain, positioned Fugu as a more reliable option for enterprise workflows than any single AI model provider in the wake of Anthropic's move on June 12 to revoke public access to its most powerful models, Claude Mythos 5 and Claude Fable 5, in the wake of a U.S. government export control order. As Ha wrote in a post today on X: "Fugu dynamically orchestrates the world’s best models to tackle complex tasks. We are proving that a well-orchestrated pool of swappable agents can match restricted frontier models like Fable and Mythos. But Fugu is about more than just performance. I believe that Orchestration Models are the next frontier, beyond bigger models. Relying on a single company’s model for national infrastructure is a massive risk. As recent export controls have shown, access to top models can disappear overnight. Collective intelligence is the practical hedge against this concentration of power. Fugu simply routes around vendor restrictions by relying on an entirely swappable agent pool." Sakana AI explicitly states that the specific models Fugu selects and how it coordinates them are proprietary, meaning this routing information is hidden from the user by design. The documentation only refers generally to a "diverse pool of powerful models," "multiple LLMs," or "specialized models" without providing a specific count. By acting as a sophisticated coordinator rather than a standalone foundation model, Fugu matches the output quality of top-tier models like Fable and Mythos on third-party benchmarks of agentic tasks, while fundamentally altering how developers deploy critical AI infrastructure. How Sakana Fugu works and where it beats Anthropic's Claude Fable 5 At its core, Sakana Fugu operates like a master general contractor. When presented with a complex request, Fugu does not attempt to execute every step itself. Instead, it breaks the problem down, delegates sub-tasks to a pool of expert foundation models, verifies their work, and synthesizes the final output. "Fugu is itself an LLM, trained to call various LLMs in an agent pool, including instances of itself recursively," the Sakana AI team noted in their technical release. Grounded in two of Sakana's 2026 research papers, TRINITY and the Conductor, the system autonomously manages the entire lifecycle of model selection and verification using learned coordination strategies rather than hand-designed workflows. To the end user, this multi-agent swarm is entirely abstracted behind a standard API endpoint. Sakana AI is offering two variants of the system to cater to different operational workloads: Fugu: A high-speed, low-latency model optimized for everyday tasks. It is designed to act as the default engine for interactive chatbots and integrates directly into coding environments like Codex. Fugu Ultra: The flagship tier engineered for complex, high-stakes tasks such as AI research, cybersecurity analysis, and multi-step patent investigations. According to Sakana, Fugu Ultra coordinates a deeper pool of experts and matches industry-leading monolithic models across rigorous scientific and reasoning benchmarks. Additionally, on the pay-as-you-go plan, standard Fugu charges a dynamic rate based on the specific underlying models activated, whereas Fugu Ultra utilizes a fixed pricing structure starting at $5 per million input tokens and $30 per million output tokens. As indicated by benchmark charts shared by Sakana, Fugu actually exceeds the performance of Anthropic's Claude Fable 5 on LiveCodeBench, an open source benchmark testing coding performance on regularly refreshed, software problem-solving tasks (Fugu Ultra: 93.2, Fugu: 92.9, Fable: 89.8), and beats the prior Claude Mythos Preview model on GPQA-D (Diamond) , a test of 198 graduate-level multiple-choice questions in biology, physics, and chemistry (Fugu Ultra: 95.5, Fugu: 95.5, Mythos Preview: 94.6). By orchestrating multiple models from different providers, Fugu essentially builds native redundancy into the AI stack. If one provider suffers an outage or faces sudden regulatory restrictions, Fugu routes around the disruption to maintain uptime. Licensing and availability Fugu is offered as a commercial, proprietary API service, not an open-source framework. Because Sakana’s core intellectual property lies in its non-obvious collaboration patterns, the specific routing information—meaning exactly which underlying models Fugu selects for a given query—remains proprietary and is intentionally hidden from the user. However, Sakana offers critical controls for enterprise data compliance. Developers can explicitly opt specific models or providers out of their Fugu routing pool to maintain strict corporate privacy standards. Additionally, users can opt out of having their prompts used for future training data. Geographically, Fugu is restricted from operating within the European Union (EU) and European Economic Area (EEA) while Sakana works to align its black-box data routing architecture with GDPR regulations. Pricing is fairly steep Fugu is available immediately in most regions—with the temporary exception of the EU and EEA—at subscription tiers and pay-as-you-go pricing. Teams can opt for monthly subscription allowances designed for individual or hands-on use: a Standard tier at $20/month for lightweight workflows, a Pro tier at $100/month providing 10x standard usage, and a Max tier at $200/month offering 20x usage for continuous, long-running tasks. I wasn't able to find the actual amount of tokens covered under these plans, but I've reached out to Ha on X for more information. As part of the initial rollout, Sakana is offering a free second month for users who subscribe to any tier by July 31, 2026. For enterprise scaling and production deployments, Sakana offers an elastic pay-as-you-go plan. Crucially for high-stakes environments, requests made under this consumption-based model are served at a higher priority than those from monthly subscription plans. Under this framework, the standard Fugu engine charges the single rate of the highest-tier underlying model involved in a query, without ever stacking multi-agent fees. The flagship Fugu Ultra tier (fugu-ultra-20260615) utilizes a fixed pricing structure per one million tokens: $5 for input, $30 for output, and $0.50 for cached input. These rates increase to $10, $45, and $1.00 respectively for extreme workloads utilizing context windows above 272K tokens. That puts it among the more expensive options compared to single AI models via provider APIs: VentureBeat Frontier AI Model API Pricing Snapshot Model Input Output Total Cost Source MiMo-V2.5 Flash $0.10 $0.30 $0.40 Xiaomi MiMo deepseek-v4-flash $0.14 $0.28 $0.42 DeepSeek deepseek-v4-pro $0.435 $0.87 $1.305 DeepSeek MiniMax-M3 $0.30 $1.20 $1.50 MiniMax Gemini 3.1 Flash-Lite $0.25 $1.50 $1.75 Google Qwen3.7-Plus $0.40 $1.60 $2.00 Alibaba Cloud MiMo-V2.5 $0.40 $2.00 $2.40 Xiaomi MiMo Grok 4.3 (low context) $1.25 $2.50 $3.75 xAI MiMo-V2.5 Pro (≤256K) $1.00 $3.00 $4.00 Xiaomi MiMo Kimi-K2.6 $0.95 $4.00 $4.95 Moonshot GLM-5.2 $1.40 $4.40 $5.80 Z.ai Grok 4.3 (high context) $2.50 $5.00 $7.50 xAI MiMo-V2.5 Pro (>256K) $2.00 $6.00 $8.00 Xiaomi MiMo Qwen3.7-Max $2.50 $7.50 $10.00 Alibaba Cloud Gemini 3.5 Flash $1.50 $9.00 $10.50 Google Gemini 3.1 Pro Preview (≤200K) $2.00 $12.00 $14.00 Google GPT-5.4 $2.50 $15.00 $17.50 OpenAI Gemini 3.1 Pro Preview (>200K) $4.00 $18.00 $22.00 Google Claude Opus 4.8 $5.00 $25.00 $30.00 Anthropic GPT-5.5 $5.00 $30.00 $35.00 OpenAI Sakana Fugu Ultra $5.00 $30.00 $35.00 Sakana AI Claude Fable 5 / Claude Mythos 5 $10.00 $50.00 $60.00 Anthropic Developers modeling operational costs should also note a significant architectural caveat in how Fugu bills for its multi-agent capabilities. According to the developer documentation, Fugu Ultra’s API responses include detailed usage fields that separate user-visible token generation from internal orchestration work. The background tokens consumed and generated when Fugu delegates sub-tasks, verifies code, or routes between underlying agents are not absorbed by the provider; they represent real token usage and are counted toward the final price of the request at standard rates. The Orchestration landscape: Fugu vs. The Field and notable benchmark performance To understand Fugu’s position in the mid-2026 AI ecosystem, it is critical to distinguish between model routing and multi-agent orchestration. Over the past year, enterprise adoption of standard routing platforms—such as Not Diamond, Martian, and the open-source RouteLLM framework—has skyrocketed. These systems act as intelligent air traffic controllers; using semantic classifiers or meta-models, they analyze an incoming prompt and predict which single foundation model will yield the highest quality or most cost-effective response, dispatching the query accordingly. Fugu operates on a fundamentally different paradigm. Rather than making a one-shot routing decision, Fugu aligns more closely with complex multi-round systems like Router-R1 (a framework introduced at NeurIPS 2025). It breaks a query down, interleaves reasoning with delegation, and dynamically assigns sub-tasks to multiple models in parallel or sequence before synthesizing a final output. While frameworks like LangGraph, CrewAI, and Microsoft AutoGen offer developers the tools to build similar multi-agent systems, they require immense manual configuration—defining roles, setting up conditional edges, and managing state across long-running loops. Fugu abstracts this operational overhead entirely. It is essentially a LangGraph-style workflow packaged as a single, black-box API endpoint. An orchestration system is ultimately bounded by the raw capabilities of the underlying models in its pool, a reality reflected in Sakana’s own benchmark testing against standalone frontier models. On rigorous coding and agentic tasks, collective intelligence shows a distinct advantage over standard models. Fugu Ultra posted a 73.7 on SWE-Bench Pro, significantly outperforming Anthropic's Claude Opus 4.8 (69.2) and OpenAI's GPT-5.5 (58.6). However, Fugu is not a silver bullet, and its performance is not a clean sweep across the board. When compared to highly specialized or restricted-access monolithic models, Fugu occasionally trails: SWE-Bench Pro: While Fugu Ultra (73.7) beat most accessible models, it was comfortably eclipsed by Anthropic’s limited-access Fable 5 (80.0), which is currently absent from Fugu's swappable pool due to the U.S. government's export control order and Anthropic's subsequent response to remove the model entirely from global usage. Humanity's Last Exam: Fugu Ultra (50.0) narrowly edged out Opus 4.8 (49.8), but again fell short of Fable 5 (53.3). Long-Context and Security: On the MRCRv2 long-context-recall test, OpenAI's GPT-5.5 maintained the lead (94.8 vs Fugu Ultra's 93.6), and Opus 4.8 remained the top performer on the CTI-REALM cybersecurity benchmark (69.6 vs Fugu Ultra's 69.4). The quantitative data points to a clear conclusion: Fugu is highly effective at boosting performance on messy, multi-step tasks (like writing a complex HTML5 game from scratch) by leaning on the combined strengths of multiple mid-tier and high-tier models. However, for sheer brute-force reasoning within a single, highly constrained domain, the industry's largest standalone models still hold the edge—provided an enterprise can maintain uninterrupted access to them. Background on Sakana's formation and noteworthy achievements to date Sakana AI was formed in Tokyo in 2023 by Llion Jones, a co-author of Google’s foundational 2017 "Attention Is All You Need" paper, and David Ha, the former head of research at Stability AI. Disillusioned by large tech company bureaucracy and the industry's hyper-fixation on scaling single, massive foundational models, the founders built Sakana around principles of biomimicry and evolutionary computing. The company's name, derived from the Japanese word for fish, reflects its core technical thesis: utilizing collective "swarm" intelligence rather than brute-force compute. Following a $2.6 billion Series B valuation in late 2025 and the recent June 2026 launch of Marlin—an autonomous, eight-hour research agent for the B2B sector—Fugu represents the commercialization of Sakana's multi-agent routing technology for everyday developers. A mixed reception among the broader AI community online The developer community has responded to Fugu by rigorously testing its practical tradeoffs, weighing its routing efficiencies against the sheer power of monolithic foundation models. AI observer, developer and influencer Chris (@ChrissGPT on X) highlighted the specific utility of Fugu over raw foundational AI. "For a single clean prompt, you probably would [use Fable 5, Mythos, or GPT-5.5 directly]," he noted, but argued that Fugu's true value emerges in messy, multi-step environments. "...whether it involves delegation, verification, synthesis, code review, research loops, security analysis... the more it would make sense to use this," he wrote. Chris also pointed out the strategic geopolitical advantage of Fugu's architecture, noting that if frontier AI access is abruptly revoked due to regulation or export controls, an orchestrator can dynamically swap models to prevent a total system failure. Creative agency owner Mark Santos (@markksantos) of Mark Studios provided a direct, real-world comparison by tasking both Fugu Ultra and Claude Opus 4.8 with building a "Crossy Road" game clone using Three.js. The results underscored the operational differences between an orchestrator and a monolithic giant: Sakana Fugu Ultra: Completed the task in 22 minutes using ~89,000 tokens for roughly $7.32. However, the final game suffered from minor logic errors, such as inverted directional turns and wonky camera angles. Claude Opus 4.8: Took 79 minutes, burned ~940,000 tokens for nearly $37.85, and got stuck in a retry loop requiring human intervention. Despite the inefficiency, it ultimately produced superior application design and functionality. Santos concluded the experiment by stating, "In terms of application functionality, quality, and design, Opus won. In terms of model speed and performance, Fugu... won". Elie Bakouch, a research engineer at cloud-based, open AI infrastructure and systems provider Prime Intellect, pointed out on X that "to be clear, this is a closed source orchestrator on top of closed source models. if before you didn't control the models, now you don't even control which ones are used or how much. this is not 'AI sovereignty'..." These early tests and reactions mirror the sentiment summarized by Reddit user GreedyWorking1499 in initial platform discussions: "Until proven otherwise, this is just a highly advanced router/wrapper, not a fundamental not a fundamental leap in intelligence like Mythos/Fable was." Yet, as enterprises increasingly demand fail-safes against single-vendor reliance, Sakana is proving that packaging collective intelligence into a single API endpoint is a highly viable commercial path.
Oracle Cut 21,000 Jobs in 12 Months, Says AI Replaced Some Roles
Oracle Corp. reduced its workforce by 21,000 employees in the past 12 months, a wider scale than previously known, including those whose jobs were eliminated by the use of artificial intelligence.
Human Capital, AI, and Labor Commoditization
Research on Upwork shows that the arrival of ChatGPT increased the importance of price in labor-market pricing while reducing the weight of credentials and reputation.
AI models capable of devastating attacks on governments and business months away, rare Five Eyes statement warns
Signal agencies in Australia, the US, the UK, New Zealand and Canada sound alarm after Trump blocks foreign nationals from Anthropic’s Fable AI model Powerful AI models capable of devastating new cyber attacks on governments and businesses are mere months away, intelligence agencies for the Five Eyes have warned in a rare joint statement, urging leaders to “act now”. The surprising public intervention by signals agencies for Australia, the US, the UK, New Zealand and Canada comes after the Trump administration earlier this month decided to block “foreign nationals” from using a much-hyped AI model built by tech company Anthropic, called Fable. Continue reading...
Chevron moves into power production with Microsoft AI deal
Company signs 20-year agreement to develop data centre in heart of US oil country that could include gas-fired plant
AI hit the memory wall — now it needs a new context tier
Presented by Solidigm As inference workloads evolve from discrete question-and-answer exchanges into persistent, multi-step agentic systems, GPU availability is no longer the most critical AI bottleneck. Instead, the bottleneck has migrated from compute to context, says Jeff Harthorn, AI applied research lead at Solidigm. "Why context management has become a primary bottleneck, more than GPU availability or compute efficiency, is the question of 2026," says Harthorn. "GPUs have gotten dramatically cheaper per FLOP. Model architectures and inference serving engines have all gotten much more efficient. But the thing that's grown faster than both of those is context. The persistent state that has to live between sessions has grown even faster than context itself." It's happening as context windows grow dramatically, making individual inputs far larger than before. Agentic AI systems chain dozens or hundreds of model calls together, each generating state that must be tracked, and enterprises are requiring that inference state persist across sessions for audit, governance, and reuse. These trends compound each other, pushing context volumes beyond what any existing memory tier was designed to handle. "Those three things are all happening at the same time, all of which are pushing context data and context memory into the stratosphere much more quickly than we're used to seeing," adds Ace Stryker, director of AI and ecosystem marketing at Solidigm. The solution is a dedicated context tier emerging between GPU memory and bulk network storage: a layer of high-performance, high-density flash designed specifically to hold and serve Key-value (KV) cache, the inference data that allows models to retain and reuse context, and retrieval data at inference speed. Nvidia has formalized this architecture under the term CMX. Storage companies including Solidigm are building SSD products optimized for this workload. "Storage has not been the first thing folks have thought about when they've been planning their enterprise infrastructure buildout," Stryker says. "In a lot of ways, it was a relatively small cost compared to compute, and it was a commodity. You just shopped around for the lowest dollar per gigabyte and called it good. But now, if your storage is not up to snuff, your ROI suffers, and it directly impacts your bottom line.” Why AI inference requires a different storage architecture than training The storage architecture that AI systems rely on today was largely inherited from training workflows. Training is sequential and write-dominated, with data moving in large blocks to and from bulk object storage. The tier structure, with high-bandwidth memory on the GPU, fast NVMe in the server, and bulk storage over the network, serves that use case reasonably well. However, inference is a different animal. Its I/O signature is fine-grained, latency-sensitive, and increasingly stateful. KV cache data and retrieval data each have distinct access patterns, but both need to be served quickly and reused across interactions. Neither fits cleanly within GPU high-bandwidth memory, which is expensive and physically constrained, nor within traditional bulk storage, which was never designed for active inference workloads. "The architectural gap that's interesting to me right now isn't at the top of the stack or the bottom, it's right in the middle," Harthon says. "A lot of what sits below the GPU HBM is being asked to do things it wasn't really designed for, which is where the most interesting systems work today is happening." One of the most visible symptoms of this gap is recomputation. In inference, the pre-fill stage processes all of the context relevant to a given session before token generation can begin. When KV cache state isn't available in a fast, accessible tier, the system recomputes it — burning GPU cycles that produce no new value. "A meaningful share of GPU cycles end up going to re-pre-filling," Harthon explains. "During all of that calculated context, that's potentially compute that's being spent reproducing state, rather than doing new work. When you start looking at the problem that way, GPU utilization starts looking like it's partly a storage problem." This reframing is driving renewed interest in a metric borrowed from networking: goodput, or useful tokens per dollar, rather than raw tokens per dollar. The AI context memory tier and how it works The industry's response is taking structural form. A new tier is emerging between GPU memory and traditional network storage, designed specifically to hold and serve inference context, a layer distinct from drives inside GPU servers (G3) and storage servers over the network (G4), engineered to serve context data back to accelerators as rapidly as possible. "If you're building a data center starting in the second half of this year, or the beginning of next year, you can't think about storage only living in two places," Stryker says. "Storage has to live in at least three places to handle the context memory tier, and that's likely to be a permanent fixture in how the infrastructure gets built going forward." It's analogous to the emergence of object storage as a category, which didn't exist until enough workloads needed it. And once it did, it developed its own primitives, SLAs, cost models, and an ecosystem of vendors. "The context tier looks like it might be on a similar arc," Harthorn says. "That volumetric pressure is causing the category to form, rather than any one vendor's road map." For infrastructure leaders, this means actively planning for the new tier rather than treating it as optional. Deploying additional NAND at this layer reduces dependency on DRAM, which is orders of magnitude more expensive per gigabyte and constrained in both availability and thermal headroom. "In terms of your investment effectiveness, you're laying out less cash to do it if you rely on the SSD layer in the way that Nvidia is now recommending and prescribing for a lot of use cases," Stryker adds. What flash needs to deliver to support AI inference Participating meaningfully in the inference stack places new demands on SSD technology. Tail latency, the worst-case performance of a drive, must be predictable, not just fast on average. An orchestration system that allocates GPU resources based on expected storage response times cannot tolerate unexpected multi-second delays. Consistent, observable performance matters more here than peak throughput. Beyond latency, density becomes a critical concern, especially at hyperscale. In data centers where power, not cost, is the binding constraint, watts per petabyte becomes the operative metric. Floating gate NAND, the manufacturing approach at the core of Solidigm's products, is suited to that calculation. Network integration via NVMe over Fabrics, RDMA, and eventual CXL support is also essential, given the tight latency budgets of active inference pipelines. "The drives have to have reliable performance characteristics, beyond the throughput side and being able to transfer as much data as possible as fast as possible, the way that training needed," Harthon says. "Now it's about being able to do it very consistently, in a way that's very observable to the people operating and orchestrating these systems." How enterprise AI leaders should plan for the context tier The standards, software primitives, and best practices being established now will define how AI inference infrastructure operates for years to come. Solidigm is engaged in that process through standards bodies, partner lab collaborations, and published research, which is critical precisely because the category is still forming. "The interesting question for the next couple of years isn't whether AI infrastructure needs more compute," Harthorn says. "It's whether it can use what it has more efficiently. A lot of that answer runs through this tier that is being built today." Sponsored articles are content produced by a company that is either paying for the post or has a business relationship with VentureBeat, and they’re always clearly marked. For more information, contact sales@venturebeat.com.
Economics & Markets
AI Shifts Payments Innovation From Features to Foundations | PYMNTS.com
Spreedly’s Judd Howard shares why competitive advantage in payments is shifting toward data, relationships and trust.
OpenAI pitches ChatGPT ads to Cannes marketers ahead of IPO
Lossmaking AI group is presenting at Cannes Lions advertising conference for the first time
Getty Images shows the inimitable value of an OpenAI photobomb
Having fought vigorously to defend its copyright, the picture agency is trying a new approach
OpenAI’s new ‘super app’ boss hopes to persuade users and potential IPO investors that the company is about way more than just chat
Folding ChatGPT and Codex into a single platform is about winning over customers—and investors.
The agentic advertising economy: From attention to action
McKinsey examines how agentic AI could shift advertising from attention capture to delegated consumer action. The piece explores scenarios involving autonomous delegation and curated ecosystems.
India’s Biggest Equity Fund Makes Contrarian IT Stocks Bet
India’s largest actively managed equity fund is betting against investor fears over artificial intelligence’s impact on the outsourcing industry as it snaps up the nation’s beaten-down IT stocks.
Qualcomm Nears Deal for AI Chip Startup Modular
Qualcomm Inc. is in advanced talks to acquire Modular Inc. in a transaction valuing the artificial intelligence infrastructure software company at about $4 billion, according to people familiar with the matter.
Internet Stocks Will Be Next Area of AI Cycle, Standard Chartered Says
Standard Chartered's Daniel Lam says the US equities market is "probably going to have a little bit of mild correction, not a lot, or stay around the current levels." He also tells Bloomberg Television that "the next area of this AI cycle is going to be the internet stocks." (Source: Bloomberg)
Sector Snapshot: Robotics Startups On Fire As Venture Funding Surges To Record Numbers In 2026
Globally, robotics startups have so far raised $18.8 billion in 2026, compared to $15 billion in the full year of 2025. The figure also handily surpasses the $14.1 billion raised in the peak venture funding year of 2021, and we still have more than six months of fundraising left.
Tencent Is Said to Mull Exits From Game Studios Like Marvelous
Tencent Holdings Ltd. is negotiating exits from several game studio investments in Japan, including Tokyo-traded Marvelous Inc., as part of a reassessment of the company’s global portfolio, people with knowledge of the matter said.
Column: Physical AI shifts from feature robots to smart robots
Over the past decade, annual venture capital invested in physical AI and robotics startups has surged from a few hundred million US dollars to nearly US$25 billion, more than a 10x increase concentrated in recent years.
European FinTech funding projections for 2026 took a hit after a 55% YoY drop in large deals
Key European FinTech investment stats in Q1 2026: European FinTech funding dropped by 31% YoY in Q1 Trend analysis showed a projected hit in funding for 2026 driven by a 55% drop in large deals (over $100m) Alan, a digital health insurance platform serving employees, freelancers, and retirees, ...
FinancialContent - Semtech and onsemi Shares Are Soaring, What You Need To Know
So this was not a broad tech rally, ... and AI-hardware supply chain specifically. Investors are distinguishing sharply between the companies building AI infrastructure, which they are rewarding, and the software and platform companies they fear AI will disrupt, which they are selling. The chip rally and the software selloff happening on the same day are two sides of the same thesis. The stock market overreacts ...
Bain & Company report identifies four imperatives shaping the future of sovereign wealth funds as global assets approach $30trln
New Bain & Company research reveals how the world’s largest state investors are recalibrating capital deployment, accelerating AI adoption, and transforming operating models for the decade ahead
Nvidia NVDA Stock Heads Below $200 as Costs, Competition, and Politics Weigh - Forex News by FX Leaders
Nvidia’s retreat from recent highs reflects mounting skepticism that even industry-leading innovation can sustain its valuation amid rising costs, geopolitical friction, and fading AI euphoria.
The AI Arms Race Isn’t About Technology – It’s About Electricity
The bottleneck has shifted from chips to power, and one under-the-radar stock just signed a 15-year, $2.6 billion AI lease that proves the trade is real. The market hasn’t priced it yet.
SpaceX sheds $400bn in market value as debut rally hits reverse
Shares in Elon Musk’s AI and rockets group tumble more than 16% following fresh rise in US bond yields
Google invests $75M in A24 for AI film partnership
We cannot provide a description for this page right now
Intellectia
Investment Trends: Goldman Sachs anticipates that AI infrastructure investments will reach $765 billion in 2023, primarily focused on critical bottlenecks such as power, memory, and networking solutions, reflecting companies' commitment to AI technology and confidence in future growth. ... Surging Power Demand: The International Energy Agency forecasts that total power consumption ...
SMH ETF flows point to AI chip trade moving away from Nvidia
VanEck Semiconductor ETF (SMH) pulled in $6.93 billion in one-day net flows—nearly a third of the week’s record tech fund inflows—as investors shifted from leveraged
VanEck's semiconductor ETF rallies as investors pile into AI hardware exposure amid strong data center demand
SMH's assets surged to $84.54 billion with an 83.2% year-to-date return as hyperscalers deploy record capital on AI infrastructure.
Info Edge’s AI startup bets cross 2x return as portfolio value reaches ₹1,268 crore - Storyboard18
The company said it has invested over ₹1,000 crore in 54 AI and deeptech startups since 2020, with consumer technology remaining its largest value creator.
NextEra Energy (NEE) Stock Could Be 7.4% Undervalued on AI Power Demand Narrative - Simply Wall St News
NextEra Energy (NEE) is back in focus after recent trading left the stock at $86.75, with returns mixed across different time frames. This has investors reassessing how its utilities and clean energy operations fit into portfolios. See our latest analysis for NextEra Energy.
Google Invests $75 Million in A24 to Develop AI-Powered Filmmaking Tools
Google has partnered with film studio A24, investing $75 million to advance the development of AI-driven tools for the film industry.
Micron Leads AI Trade Higher. Expectations Are Rising Ahead of the Memory Chipmaker's Earnings.
Micron Technology led other AI hardware stocks higher to start the week, ahead of the memory chipmaker's earnings.
AI in Medical Imaging Market to Reach $8.56 Billion by 2030 at 30% CAGR | GE Healthcare, Butterfly Network, Enlitic Among Key Players
The global AI in medical imaging market valued at 1 75 billion in 2024 will reach 8 56 billion by 2030 expanding at a compound annual growth rate CAGR of 30 over the forecast period The scale of that growth ...
South Korean chipmakers are being paid such massive bonuses it’s becoming an inflation problem for the central bank
The Bank of Korea is worried that a wage-price spiral is increasing consumer demand and pushing up prices.
AI Employment Impact: AI boom's US employment, wage impact muted so far, ECB study finds, ETHRWorld
AI Employment Impact: Firms have been investing heavily in AI in recent years, raising fears that humans will be replaced at increasing rates, curbing overall employment and widening inequality along the way.
Is China Closing the A.I. Gap Faster Than Expected?
Silicon Valley and corporate America are increasingly turning to cheaper, open-source artificial intelligence models built in China.
Bain tests software takeover targets by vibecoding AI replicas
Private equity groups swiftly recreate software products to gauge their competitive advantages
Gas Stations Accused of Using AI to Boost California Prices
A group of California consumers claimed in a lawsuit that gas station owners including Walmart Inc., Marathon Petroleum Corp., BP Plc and 7-Eleven Inc. are using artificial intelligence to illegally manipulate pump prices in the state that already has the highest rates in the US.
Alibaba's AI video model rises to No. 2 in global rankings, as OpenAI's Sora and ByteDance's Seedance fall away
Alibaba Cloud on Sunday released HappyHorse 1.1, a major upgrade to its AI video generation model that the company says delivers production-ready video synthesis across core content creation scenarios. The model is now live on Alibaba Cloud Model Studio with full API access for enterprise customers and developers, accompanied by a 40% sitewide launch discount for the first two weeks. The release arrives at a moment of remarkable upheaval in the AI video generation market — and Alibaba appears keenly aware of the timing. OpenAI discontinued Sora after it proved financially unsustainable. ByteDance indefinitely shelved the international rollout of Seedance 2.0 following a barrage of copyright complaints from Hollywood studios. For enterprise procurement teams that had been evaluating or integrating those tools into marketing, advertising, and content production workflows, the competitive landscape has contracted sharply in a matter of months. That contraction creates both an opportunity and a test for Alibaba. HappyHorse 1.1 is not a research demo or a consumer toy — it is an API-first product built for integration into enterprise software stacks, priced for volume, and backed by a $52.7 billion global infrastructure buildout. Whether it can convert technical capability into enterprise adoption, particularly in Western markets navigating intensifying U.S.-China tech tensions, will determine whether Alibaba can establish itself as a serious player in the generative video market that analysts expect to reach tens of billions of dollars by the end of the decade. How HappyHorse climbed from anonymous benchmark entry to top-ranked video model HappyHorse first appeared in early April as an anonymous submission on the Artificial Analysis Video Arena, an independent benchmarking platform where real users compare model outputs in blind, side-by-side evaluations. The model immediately claimed the top position in both text-to-video and image-to-video rankings. Alibaba was subsequently confirmed as the creator, revealing it was built by the company's ATH (Alibaba Token Hub) AI Innovation Unit — a team previously part of the Future Life Lab under the Taobao and Tmall Group before a strategic organizational restructuring. According to Arena.ai, HappyHorse 1.0 now holds the No. 2 position across all three Video Arena leaderboards. The platform noted the model scores 1,444 in both text-to-video and image-to-video categories, leading Google's Veo-3.1 (with audio) by 69 points in text-to-video and xAI's Grok-Imagine-Video by 23 points in image-to-video. In Elo-based ranking systems like Arena's, models gain or lose points based on whether users prefer their outputs in head-to-head comparisons, meaning persistent double-digit leads reflect a consistent quality gap as perceived by human evaluators — not a statistical fluke. The model's architecture helps explain why. According to community-compiled technical documentation, HappyHorse is built around a 15-billion-parameter unified self-attention Transformer that processes text, image, video, and audio tokens within a single token sequence. Unlike many competitors that stitch together separate models for video and audio, HappyHorse operates as a unified system that handles all modalities in a single generation pass, eliminating the need for third-party dubbing or post-processing audio tools. For enterprise buyers evaluating total cost of ownership, that architectural simplicity translates directly into fewer integration points, fewer vendor dependencies, and faster time to production. What the 1.1 upgrade fixes — and why it matters for commercial video production The 1.1 upgrade targets a set of pain points that enterprise video production teams know intimately. Alibaba Cloud described the release as "systematically optimized across core content generation scenarios," and the specific improvements reveal a model that has been tuned for commercial deployment rather than viral social media demos. The most consequential upgrade is multi-image reference capability, which Alibaba calls R2V (Reference-to-Video). The feature allows users to upload multiple character reference images and maintain consistent identity across generated video — directly addressing one of the hardest problems in AI video production, where subjects tend to drift in appearance between frames or shots. For brands producing advertising campaigns, product videos, or serialized marketing content, identity consistency is not a nice-to-have; it is a requirement that has historically forced teams back to traditional production methods. Motion quality receives a significant overhaul, with what Alibaba describes as "strengthened motion modeling" that addresses prior limitations in speed and fluidity. The company also made targeted improvements to visual texture, specifically calling out the elimination of "facial oiliness," "over-sharpening," and "unnatural textures" — artifacts that have plagued commercial AI video since the technology emerged and that immediately signal to viewers that content is machine-generated. Two additional upgrades round out the release. HappyHorse 1.1 improves audio-visual synchronization, including what Alibaba claims is "zero-drift lip sync" for dialogue scenes and context-aware speech pacing — building on the 1.0 version's already notable ability to generate up to 15 seconds of 1080p video with synchronized audio output. The model also improves instruction-following for long and complex prompts, a critical differentiator for enterprise users who need to specify precise camera movements, lighting conditions, and narrative beats in a single generation pass rather than iterating through dozens of attempts. Sora's collapse and Seedance's freeze leave enterprise buyers with fewer choices than ever The competitive context surrounding this launch is unusually favorable for Alibaba, and it is worth understanding why. OpenAI's Sora web and app experiences were discontinued on April 26, with the Sora API set to follow on September 24. The shutdown came after the product proved financially untenable: Sora cost roughly $1 million per day to operate but generated only about $2.1 million in total revenue, while active users dropped from a peak near 1 million to under 500,000. For enterprise teams that had integrated Sora into production pipelines, the abrupt withdrawal underscored the risks of depending on AI products that lack a sustainable business model — a cautionary tale that procurement officers are unlikely to forget quickly. ByteDance's Seedance 2.0, which many considered Sora's most formidable successor, ran into a different kind of wall. Netflix, Warner Bros., Disney, Paramount, and Sony sent legal threats to ByteDance over allegations of systematic copyright infringement after users generated viral clips featuring Hollywood intellectual property. ByteDance indefinitely postponed the international launch, and the global rollout remains suspended. That leaves Google's Veo 3.1 as the primary Western competitor in the enterprise video generation space. But Alibaba's Arena rankings suggest HappyHorse is outperforming Veo on user-perceived quality, and the 40% launch discount on Alibaba Cloud Model Studio could make HappyHorse significantly cheaper at scale. At the 1.0 level, pricing through third-party API platforms ran roughly $1.82 per 10-second clip at 720p and $3.12 at 1080p. With the promotional pricing, HappyHorse 1.1 could bring production-quality AI video generation within reach of mid-market companies and agencies that previously considered the technology too expensive for anything beyond experimentation. Alibaba's $52.7 billion infrastructure bet gives HappyHorse a distribution advantage rivals can't match HappyHorse 1.1 does not exist in isolation. It sits atop a global infrastructure offensive that distinguishes Alibaba from pure-play AI model companies that build impressive technology but lack the physical and commercial machinery to serve regulated enterprise customers at scale. Just five days before the HappyHorse 1.1 launch, Alibaba Cloud opened its first data centers in France, establishing its third European hub after Germany and the United Kingdom. The Paris region features two availability zones, bringing the company's global footprint to 105 availability zones across 32 regions. "The expansion of our cloud infrastructure into France reinforces our ongoing commitment to empowering European businesses with sovereign, secure, and intelligent solutions," said Dr. Feifei Li, Alibaba Cloud's CTO and president of international business, in the company's announcement. In Japan, the company opened its fifth data center in Tokyo on June 19. As reported by Data Center Dynamics, CEO Eddie Wu has committed to investing $52.7 billion in building a "unified global cloud network," with the company later considering increasing this to $69 billion. This year alone, Alibaba has launched new regions in Mexico, Thailand, Malaysia's Johor, and France. The France deployment is also part of Alibaba Cloud's plan to roll out enterprise-grade agentic AI services across Europe in the second half of the year, including AgentRun (a development platform for AI agents), STAROps (an intelligent operations platform), and ACS Agent Sandbox (which provides hardware-level security isolation for agent workloads). The infrastructure buildout serves a dual purpose for a product like HappyHorse. Running a 15-billion-parameter video generation model with integrated audio is extraordinarily compute-intensive, and having local infrastructure reduces latency for enterprise API calls while keeping customer data within regulatory boundaries. For European buyers operating under the European Commission's new tech sovereignty framework — published June 3 with the explicit goal of protecting the bloc's "digital independence" — the ability to run AI video generation workloads on locally hosted infrastructure is not a luxury. It is increasingly a compliance requirement. The Pentagon listing and geopolitical risk loom over Alibaba's Western ambitions Alibaba's global push is unfolding under significant geopolitical headwinds that enterprise buyers cannot afford to ignore. The Pentagon added Alibaba, along with BYD and Baidu, to its list of Chinese military companies on June 8, preventing them from securing U.S. defense contracts. Alibaba rejected the designation, saying it is "not a Chinese military company nor part of any military-civil fusion strategy." The listing does not automatically trigger sanctions, and it does not directly restrict commercial transactions between private U.S. companies and Alibaba. But it adds a layer of reputational and regulatory complexity to procurement decisions, particularly for companies with U.S. government exposure, defense supply chain connections, or transatlantic operations. Enterprise technology purchases are rarely evaluated on technical merit alone — vendor risk assessments, board-level compliance reviews, and geopolitical scenario planning all factor into buying decisions for cloud infrastructure and AI tooling. For European customers specifically, the calculus is layered in a different way. The continent's growing emphasis on digital sovereignty cuts in two directions simultaneously: it creates demand for alternatives to the dominant U.S. hyperscalers (Amazon Web Services, Microsoft Azure, and Google Cloud control roughly 70 percent of European cloud infrastructure revenue, according to Synergy Research Group), but it also raises questions about whether a Chinese provider represents a meaningful improvement in strategic autonomy. Alibaba's strategy of building sovereignty-compliant infrastructure in-market is a direct attempt to answer that question — but the Pentagon listing ensures it will be asked repeatedly. What enterprise teams should watch as the AI video market consolidates The practical implications of HappyHorse 1.1 for enterprise teams are substantial. HappyHorse supports four modes of generation — text-to-video, image-to-video, subject-to-video, and the newly added video editing — covering the full spectrum of commercial video needs from ideation through production to post-production, all with integrated audio at no additional cost. That breadth of capability, delivered through a single API endpoint, simplifies what has historically been a fragmented and expensive production pipeline. The question going forward is whether Alibaba can convert benchmark dominance and competitive timing into durable enterprise relationships. The company plans to release HappyHorse through Alibaba Cloud Model Studio with full enterprise SLAs, security certifications, and regional compliance — the table stakes that separate research breakthroughs from production-grade services. Watch for customer disclosures, usage metrics, and whether third-party platforms like fal.ai and Atlas Cloud (which already host HappyHorse 1.0) update to the 1.1 version quickly, which would signal genuine developer demand beyond Alibaba's own ecosystem. The AI video generation market entered 2026 with three credible enterprise contenders. One is dead. One is frozen. And the one still standing is a Chinese company backed by $52.7 billion in infrastructure spending, ranked No. 2 across every major independent benchmark, and offering a 40% discount to anyone willing to place the bet. In enterprise technology, the best product does not always win — but it rarely loses when the competition has already left the field.
Marathon, BP, other gas stations accused of AI-powered scheme to hike gas prices in US
A lawsuit alleges that major gas station chains conspired to use an AI-powered tool to artificially inflate gasoline prices in California.
Gas stations accused of algorithmic collusion to raise prices in California
Marathon Petroleum, 7-Eleven, and other retailers are accused of using an AI-powered tool from Kalibrate to artificially inflate gasoline prices in California.
Tencent Tests AI Assistant Xiaowei in WeChat
Tencent is internally testing Xiaowei, an AI assistant, within WeChat in China to bolster its AI capabilities and compete with ByteDance and Alibaba.
Nvidia Stock Steady Near $210 as Robotics, Supercomputer Announcements Roll Out
Nvidia's stock remains stable amid new AI partnerships and product announcements, despite recent volatility and competition from customers developing their own AI chips.
No Claude Fable 5? No problem: Sakana achieves frontier performance with new Fugu multi-model, auto synthesis system
Last night, the increasingly enterprise-focused AI startup Sakana launched Fugu, a multi-agent orchestration system that delivers frontier-level AI performance through a single, OpenAI-compatible API. Designed for developers, enterprises, and nations seeking resilience against vendor lock-in and geopolitical export controls, Fugu (Japanese for "pufferfish"), bypasses the traditional monolithic model structure by dynamically routing queries to a swappable pool of specialized AI agents. Sakana CEO and co-founder David Ha, formerly of Google Brain, positioned Fugu as a more reliable option for enterprise workflows than any single AI model provider in the wake of Anthropic's move on June 12 to revoke public access to its most powerful models, Claude Mythos 5 and Claude Fable 5, in the wake of a U.S. government export control order. As Ha wrote in a post today on X: "Fugu dynamically orchestrates the world’s best models to tackle complex tasks. We are proving that a well-orchestrated pool of swappable agents can match restricted frontier models like Fable and Mythos. But Fugu is about more than just performance. I believe that Orchestration Models are the next frontier, beyond bigger models. Relying on a single company’s model for national infrastructure is a massive risk. As recent export controls have shown, access to top models can disappear overnight. Collective intelligence is the practical hedge against this concentration of power. Fugu simply routes around vendor restrictions by relying on an entirely swappable agent pool." Sakana AI explicitly states that the specific models Fugu selects and how it coordinates them are proprietary, meaning this routing information is hidden from the user by design. The documentation only refers generally to a "diverse pool of powerful models," "multiple LLMs," or "specialized models" without providing a specific count. By acting as a sophisticated coordinator rather than a standalone foundation model, Fugu matches the output quality of top-tier models like Fable and Mythos on third-party benchmarks of agentic tasks, while fundamentally altering how developers deploy critical AI infrastructure. How Sakana Fugu works and where it beats Anthropic's Claude Fable 5 At its core, Sakana Fugu operates like a master general contractor. When presented with a complex request, Fugu does not attempt to execute every step itself. Instead, it breaks the problem down, delegates sub-tasks to a pool of expert foundation models, verifies their work, and synthesizes the final output. "Fugu is itself an LLM, trained to call various LLMs in an agent pool, including instances of itself recursively," the Sakana AI team noted in their technical release. Grounded in two of Sakana's 2026 research papers, TRINITY and the Conductor, the system autonomously manages the entire lifecycle of model selection and verification using learned coordination strategies rather than hand-designed workflows. To the end user, this multi-agent swarm is entirely abstracted behind a standard API endpoint. Sakana AI is offering two variants of the system to cater to different operational workloads: Fugu: A high-speed, low-latency model optimized for everyday tasks. It is designed to act as the default engine for interactive chatbots and integrates directly into coding environments like Codex. Fugu Ultra: The flagship tier engineered for complex, high-stakes tasks such as AI research, cybersecurity analysis, and multi-step patent investigations. According to Sakana, Fugu Ultra coordinates a deeper pool of experts and matches industry-leading monolithic models across rigorous scientific and reasoning benchmarks. Additionally, on the pay-as-you-go plan, standard Fugu charges a dynamic rate based on the specific underlying models activated, whereas Fugu Ultra utilizes a fixed pricing structure starting at $5 per million input tokens and $30 per million output tokens. As indicated by benchmark charts shared by Sakana, Fugu actually exceeds the performance of Anthropic's Claude Fable 5 on LiveCodeBench, an open source benchmark testing coding performance on regularly refreshed, software problem-solving tasks (Fugu Ultra: 93.2, Fugu: 92.9, Fable: 89.8), and beats the prior Claude Mythos Preview model on GPQA-D (Diamond) , a test of 198 graduate-level multiple-choice questions in biology, physics, and chemistry (Fugu Ultra: 95.5, Fugu: 95.5, Mythos Preview: 94.6). By orchestrating multiple models from different providers, Fugu essentially builds native redundancy into the AI stack. If one provider suffers an outage or faces sudden regulatory restrictions, Fugu routes around the disruption to maintain uptime. Licensing and availability Fugu is offered as a commercial, proprietary API service, not an open-source framework. Because Sakana’s core intellectual property lies in its non-obvious collaboration patterns, the specific routing information—meaning exactly which underlying models Fugu selects for a given query—remains proprietary and is intentionally hidden from the user. However, Sakana offers critical controls for enterprise data compliance. Developers can explicitly opt specific models or providers out of their Fugu routing pool to maintain strict corporate privacy standards. Additionally, users can opt out of having their prompts used for future training data. Geographically, Fugu is restricted from operating within the European Union (EU) and European Economic Area (EEA) while Sakana works to align its black-box data routing architecture with GDPR regulations. Pricing is fairly steep Fugu is available immediately in most regions—with the temporary exception of the EU and EEA—at subscription tiers and pay-as-you-go pricing. Teams can opt for monthly subscription allowances designed for individual or hands-on use: a Standard tier at $20/month for lightweight workflows, a Pro tier at $100/month providing 10x standard usage, and a Max tier at $200/month offering 20x usage for continuous, long-running tasks. I wasn't able to find the actual amount of tokens covered under these plans, but I've reached out to Ha on X for more information. As part of the initial rollout, Sakana is offering a free second month for users who subscribe to any tier by July 31, 2026. For enterprise scaling and production deployments, Sakana offers an elastic pay-as-you-go plan. Crucially for high-stakes environments, requests made under this consumption-based model are served at a higher priority than those from monthly subscription plans. Under this framework, the standard Fugu engine charges the single rate of the highest-tier underlying model involved in a query, without ever stacking multi-agent fees. The flagship Fugu Ultra tier (fugu-ultra-20260615) utilizes a fixed pricing structure per one million tokens: $5 for input, $30 for output, and $0.50 for cached input. These rates increase to $10, $45, and $1.00 respectively for extreme workloads utilizing context windows above 272K tokens. That puts it among the more expensive options compared to single AI models via provider APIs: VentureBeat Frontier AI Model API Pricing Snapshot Model Input Output Total Cost Source MiMo-V2.5 Flash $0.10 $0.30 $0.40 Xiaomi MiMo deepseek-v4-flash $0.14 $0.28 $0.42 DeepSeek deepseek-v4-pro $0.435 $0.87 $1.305 DeepSeek MiniMax-M3 $0.30 $1.20 $1.50 MiniMax Gemini 3.1 Flash-Lite $0.25 $1.50 $1.75 Google Qwen3.7-Plus $0.40 $1.60 $2.00 Alibaba Cloud MiMo-V2.5 $0.40 $2.00 $2.40 Xiaomi MiMo Grok 4.3 (low context) $1.25 $2.50 $3.75 xAI MiMo-V2.5 Pro (≤256K) $1.00 $3.00 $4.00 Xiaomi MiMo Kimi-K2.6 $0.95 $4.00 $4.95 Moonshot GLM-5.2 $1.40 $4.40 $5.80 Z.ai Grok 4.3 (high context) $2.50 $5.00 $7.50 xAI MiMo-V2.5 Pro (>256K) $2.00 $6.00 $8.00 Xiaomi MiMo Qwen3.7-Max $2.50 $7.50 $10.00 Alibaba Cloud Gemini 3.5 Flash $1.50 $9.00 $10.50 Google Gemini 3.1 Pro Preview (≤200K) $2.00 $12.00 $14.00 Google GPT-5.4 $2.50 $15.00 $17.50 OpenAI Gemini 3.1 Pro Preview (>200K) $4.00 $18.00 $22.00 Google Claude Opus 4.8 $5.00 $25.00 $30.00 Anthropic GPT-5.5 $5.00 $30.00 $35.00 OpenAI Sakana Fugu Ultra $5.00 $30.00 $35.00 Sakana AI Claude Fable 5 / Claude Mythos 5 $10.00 $50.00 $60.00 Anthropic Developers modeling operational costs should also note a significant architectural caveat in how Fugu bills for its multi-agent capabilities. According to the developer documentation, Fugu Ultra’s API responses include detailed usage fields that separate user-visible token generation from internal orchestration work. The background tokens consumed and generated when Fugu delegates sub-tasks, verifies code, or routes between underlying agents are not absorbed by the provider; they represent real token usage and are counted toward the final price of the request at standard rates. The Orchestration landscape: Fugu vs. The Field and notable benchmark performance To understand Fugu’s position in the mid-2026 AI ecosystem, it is critical to distinguish between model routing and multi-agent orchestration. Over the past year, enterprise adoption of standard routing platforms—such as Not Diamond, Martian, and the open-source RouteLLM framework—has skyrocketed. These systems act as intelligent air traffic controllers; using semantic classifiers or meta-models, they analyze an incoming prompt and predict which single foundation model will yield the highest quality or most cost-effective response, dispatching the query accordingly. Fugu operates on a fundamentally different paradigm. Rather than making a one-shot routing decision, Fugu aligns more closely with complex multi-round systems like Router-R1 (a framework introduced at NeurIPS 2025). It breaks a query down, interleaves reasoning with delegation, and dynamically assigns sub-tasks to multiple models in parallel or sequence before synthesizing a final output. While frameworks like LangGraph, CrewAI, and Microsoft AutoGen offer developers the tools to build similar multi-agent systems, they require immense manual configuration—defining roles, setting up conditional edges, and managing state across long-running loops. Fugu abstracts this operational overhead entirely. It is essentially a LangGraph-style workflow packaged as a single, black-box API endpoint. An orchestration system is ultimately bounded by the raw capabilities of the underlying models in its pool, a reality reflected in Sakana’s own benchmark testing against standalone frontier models. On rigorous coding and agentic tasks, collective intelligence shows a distinct advantage over standard models. Fugu Ultra posted a 73.7 on SWE-Bench Pro, significantly outperforming Anthropic's Claude Opus 4.8 (69.2) and OpenAI's GPT-5.5 (58.6). However, Fugu is not a silver bullet, and its performance is not a clean sweep across the board. When compared to highly specialized or restricted-access monolithic models, Fugu occasionally trails: SWE-Bench Pro: While Fugu Ultra (73.7) beat most accessible models, it was comfortably eclipsed by Anthropic’s limited-access Fable 5 (80.0), which is currently absent from Fugu's swappable pool due to the U.S. government's export control order and Anthropic's subsequent response to remove the model entirely from global usage. Humanity's Last Exam: Fugu Ultra (50.0) narrowly edged out Opus 4.8 (49.8), but again fell short of Fable 5 (53.3). Long-Context and Security: On the MRCRv2 long-context-recall test, OpenAI's GPT-5.5 maintained the lead (94.8 vs Fugu Ultra's 93.6), and Opus 4.8 remained the top performer on the CTI-REALM cybersecurity benchmark (69.6 vs Fugu Ultra's 69.4). The quantitative data points to a clear conclusion: Fugu is highly effective at boosting performance on messy, multi-step tasks (like writing a complex HTML5 game from scratch) by leaning on the combined strengths of multiple mid-tier and high-tier models. However, for sheer brute-force reasoning within a single, highly constrained domain, the industry's largest standalone models still hold the edge—provided an enterprise can maintain uninterrupted access to them. Background on Sakana's formation and noteworthy achievements to date Sakana AI was formed in Tokyo in 2023 by Llion Jones, a co-author of Google’s foundational 2017 "Attention Is All You Need" paper, and David Ha, the former head of research at Stability AI. Disillusioned by large tech company bureaucracy and the industry's hyper-fixation on scaling single, massive foundational models, the founders built Sakana around principles of biomimicry and evolutionary computing. The company's name, derived from the Japanese word for fish, reflects its core technical thesis: utilizing collective "swarm" intelligence rather than brute-force compute. Following a $2.6 billion Series B valuation in late 2025 and the recent June 2026 launch of Marlin—an autonomous, eight-hour research agent for the B2B sector—Fugu represents the commercialization of Sakana's multi-agent routing technology for everyday developers. A mixed reception among the broader AI community online The developer community has responded to Fugu by rigorously testing its practical tradeoffs, weighing its routing efficiencies against the sheer power of monolithic foundation models. AI observer, developer and influencer Chris (@ChrissGPT on X) highlighted the specific utility of Fugu over raw foundational AI. "For a single clean prompt, you probably would [use Fable 5, Mythos, or GPT-5.5 directly]," he noted, but argued that Fugu's true value emerges in messy, multi-step environments. "...whether it involves delegation, verification, synthesis, code review, research loops, security analysis... the more it would make sense to use this," he wrote. Chris also pointed out the strategic geopolitical advantage of Fugu's architecture, noting that if frontier AI access is abruptly revoked due to regulation or export controls, an orchestrator can dynamically swap models to prevent a total system failure. Creative agency owner Mark Santos (@markksantos) of Mark Studios provided a direct, real-world comparison by tasking both Fugu Ultra and Claude Opus 4.8 with building a "Crossy Road" game clone using Three.js. The results underscored the operational differences between an orchestrator and a monolithic giant: Sakana Fugu Ultra: Completed the task in 22 minutes using ~89,000 tokens for roughly $7.32. However, the final game suffered from minor logic errors, such as inverted directional turns and wonky camera angles. Claude Opus 4.8: Took 79 minutes, burned ~940,000 tokens for nearly $37.85, and got stuck in a retry loop requiring human intervention. Despite the inefficiency, it ultimately produced superior application design and functionality. Santos concluded the experiment by stating, "In terms of application functionality, quality, and design, Opus won. In terms of model speed and performance, Fugu... won". Elie Bakouch, a research engineer at cloud-based, open AI infrastructure and systems provider Prime Intellect, pointed out on X that "to be clear, this is a closed source orchestrator on top of closed source models. if before you didn't control the models, now you don't even control which ones are used or how much. this is not 'AI sovereignty'..." These early tests and reactions mirror the sentiment summarized by Reddit user GreedyWorking1499 in initial platform discussions: "Until proven otherwise, this is just a highly advanced router/wrapper, not a fundamental not a fundamental leap in intelligence like Mythos/Fable was." Yet, as enterprises increasingly demand fail-safes against single-vendor reliance, Sakana is proving that packaging collective intelligence into a single API endpoint is a highly viable commercial path.
Microsoft CEO Satya Nadella warns against concentration of AI power
Satya Nadella cautions against the risks of allowing a small number of companies to control the development and deployment of powerful AI technologies.
AI law firm wins UK court case for first time
Freelancer paid about £400 for technology to draft documents for £7,000 claim
How AI fits in the energy development workflow | Utility Dive
Power consumption is projected to hit record levels again in 2026. And the infrastructure to meet that demand is still years behind. Development teams feel this acutely. Pipelines that used to carry 20 active sites now need to hold 80 to hit the same conversion rate. Project timelines have compressed while headcount hasn't grown to match. Other industries have responded to this kind of pressure by turning to AI. The energy ...
We’re Measuring AI Productivity But Missing The Human Capacity Problem
AI is expected to increase productivity but human capacity remains limited. To fully leverage AI, humans need space for learning and deeper thinking.
The Surge of Slop
A scientific analysis shows that the explosion in Amazon e-books since 2022 is driven by AI-generated content, which now outnumbers human-written books.
75,000 AI songs a day, but Europe’s music startups are looking somewhere smarter
Two years ago, a Sónar+D panel called “Generating Panic?” asked whether AI was about to overwhelm music. By April 2026, Deezer had a number: around 75,000 fully AI-generated tracks a day, about 44% of all the new music arriving on the platform. The panel asked, the data answered. Then you look at what people play. […]
An AI venture soared as it touted links to Trump. Now it’s fighting for survival
Shares in Fermi collapsed after it failed to secure an anchor tenant and its CEO was fired. Can it ever fulfil its promise?
Alem Capital Invests in Regional Fund Backing Central Asian Tech Startups - Bloomberg
Kazakhstan’s Alem Capital Management anchored a regional venture capital fund backing technology startups across Central Asia and the Caucasus.
Seedcamp eyes physical AI with $320M raise - PitchBook
The Revolut-backer will spread the capital across its core seed strategy and a follow-on fund.
Investors are not happy about Google losing top AI talent
Concerns are mounting among investors as Google continues to see an exodus of key artificial intelligence researchers and engineers.
Labor, Society & Culture
Oracle Cut 21,000 Jobs in 12 Months, Says AI Replaced Some Roles
Oracle Corp. reduced its workforce by 21,000 employees in the past 12 months, a wider scale than previously known, including those whose jobs were eliminated by the use of artificial intelligence.
Simon Johnson: ‘Nobody needs as many white-collar workers as they used to’
The Nobel laureate and former IMF chief economist on how to prepare for what AI will do to jobs
Opinion | AI is sparking a blue-collar jobs boom. Now comes the hard part. - The Washington Post
Meeting a massive need for more electricians, welders and plumbers will take real ambition.
Human Capital, AI, and Labor Commoditization
Research on Upwork shows that the arrival of ChatGPT increased the importance of price in labor-market pricing while reducing the weight of credentials and reputation.
Christoph Ebhart - Sondero | LinkedIn
GTM Exec Dinner in Munich 🇩🇪🐞 We talked about: 1. AI We see a lot of small improvements trough AI in workflows that add up to become more efficient. There is no major breakthrough use-case yet. One peer reduced their SDR team from 15 to 1 AI enabled person and the setup works (Large Enterprise heavy).
AI and the US labour market: effects on employment growth - Pension Policy International
As firms around the world adopt AI tools, the impact of AI on labour markets is being widely discussed.[1] While AI’s potential to disrupt job markets could be significant, its effects on aggregate employment appear to be muted so far. Still, there is growing evidence that AI is negatively ...
When Workers Lose to AI | Foreign Affairs
How the U.S. government can soften the blow of automation.
The autonomous business is coming. Here's why that shift is good news for professionals | ZDNET
Companies are investing in AI agents and cutting staff, but talented professionals will find new opportunities.
Beware a two-speed workforce: AI skills are the new great divide - Entrepreneur United Kingdom - Entrepreneur United Kingdom
AI is transforming work faster than expected. Without investment in skills and training, the UK risks creating a two-speed workforce and economy.
Workers Are Emerging as the Next Big New AI Bottleneck - Linkdood Technologies
Professional services, healthcare, education, software development, customer service, manufacturing, finance, and media are expected to undergo significant transformation due to AI adoption. Not always. Productivity gains often take time because organizations must redesign workflows, train ...
AI and VR Revolutionize Hiring: Beyond Resumes to Predict Job Performance
The hiring industry is increasingly using AI, gamified assessments, and VR to predict job performance, though experts warn these should complement rather than replace human judgment.
Workday must face nationwide discrimination claims over AI tools
Workday is facing legal action regarding discrimination claims linked to its AI-powered tools.
US judge finds Workday must face Calif. discrimination claims over AI hiring tools
A US District Judge rejected Workday's attempt to dismiss claims that its AI hiring tools violated the California Fair Employment and Housing Act and the Americans with Disabilities Act.
The running list: major tech layoffs in 2026 where employers cited AI
A tracking list of significant tech industry layoffs in 2026 that have been explicitly attributed to the adoption of AI technologies.
Human skills still matter: Why crisis decision-making may save jobs from AI disruption - Digital Journal
New evidence suggests that human-centred competencies—judgement under pressure, interpersonal engagement and adaptive reasoning—will remain critical assets in the future workforce.
Employers, Workers Disconnected on AI in the Workforce
Prudential’s report shows that while 71% of employers categorize AI as a positive tool, only 51% of employees agree.
AI impact on jobs is overstated, says IBM CHRO
IBM's CHRO suggests that the narrative surrounding AI and job displacement is more complex than simple mass replacement.
Employers tap AI for benefits, but workers remain cautious
A new Prudential study finds employers embrace AI for employee benefits and the personalization it can bring, but workers remain concerned about trust and data privacy.
Huge job losses from AI could tank stocks and pave the way for basic income, short-seller Carson Block says
Sweeping job losses from AI could spark a wave of withdrawals from the stock market and make government checks a necessity, Carson Block said.
Why AI Will Redefine Independent Agency Performance
Independent insurance agents are entering the most transformative era in the history of insurance distribution. Artificial intelligence is already
The next evolution of Enterprise AI: From governance frameworks to runtime accountability - Express Computer
By Prem Brahmandam, The Hartford India Leadership Team Artificial intelligence has moved past the stage of isolated pilots. In large enterprises, it is now becoming part of the operating core. […]
Ethics of Artificial Intelligence - AI | UNESCO
UNESCO promotes ethical AI through global recommendations, guiding responsible design, development, and use of artificial intelligence.
Import AI 462: Superpersuasion; self-sustaining AI; paths to ASI
The results are definitive: across four experiments involving 18,978 conversations across 6,923 people, AI systems are, today, better than humans at text-based persuasion with real world consequences - though humans can be equivalent to them if we place some artificial constraints on the AI systems.
Artificial intelligence adoption is accelerating but so are the risks - Digital Journal
Artificial intelligence deployment must be accompanied by a commensurate investment in governance and risk management. The experiences of major organisations serve as a reminder that AI is not inherently reliable, secure or unbiased.
Top Intel Agencies Say AI-Driven Cyber Catastrophes Are Imminent: 'The Timeline Is Not Years, It Is Months'
"Breaches will occur," the intelligence alliance warns org leaders.
AI as “Arbitrary” Intelligence | The Regulatory Review
Federal Agencies’ growing use of AI raises questions about how judges can adequately review those agencies’ decisions.
The Hidden Environmental Cost of the New AI Revolution - Linkdood Technologies
Electricity often dominates discussions about AI infrastructure, but water may become an equally important issue. Servers generate enormous heat. To prevent overheating, facilities rely on cooling systems that consume significant amounts of water or energy. Many proposed developments are located in regions already facing climate pressures, drought risks, or population growth challenges...
Half of London firms report skills gap amid AI boom - BBC News
Of the just over 2,000 business leaders surveyed, half say their staff have the necessary skills.
Rep. Sam Liccardo unveils AI workforce tax credit bill - POLITICO
The legislation aims to boost training programs for the jobs needed in an AI future.
Opinion | New York City politicians push a moratorium on AI in schools - The Washington Post
Forward-thinking educators are finding creative ways to use artificial intelligence to help students. Meanwhile, the majority of New York’s City Council is pressing Mayor Zohran Mamdani (D) to “immediately pause” any use of AI in schools.
FEU Tech Partners with OpenAI to Become Philippines' First AI-Native University
FEU Tech is integrating OpenAI technology across its academic and administrative functions to prepare students for an AI-driven economy.
Technology & Infrastructure
Agentic AI: the next stage of automation or a new model of work? — Retail Technology Innovation Hub
For most of the past decade, business automation has been associated primarily with robotic process automation (RPA) and simple chatbots designed to handle repetitive customer inquiries. These technologies delivered measurable operational benefits. They accelerated processes, reduced costs, and impr
Agentic AI is scaling in manufacturing, but infrastructure gaps remain | Manufacturing Dive
Experts from the Institute of Electrical and Electronics Engineers, Iterate.ai, Altimetrik and Amtech Software weigh in on the state of agentic AI adoption in manufacturing.
OpenCode: The Terminal-First AI Coding Revolution Dominating Developer Workflows in 2026
OpenCode has become a leading model-agnostic coding agent, supporting over 75 providers and serving 7.5 million active developers as of June 2026.
Agentic coding and persistent returns to expertise
Anthropic research on Claude Code shows that agentic coding shifts humans toward planning while AI handles execution. Domain expertise and management-style instruction remain critical for productivity.
Agentic AI ROI: Measure Business Value & Productivity Impact
Agentic AI ROI is not theoretical. Early adopters are seeing measurable returns through labor productivity gains, accelerated decision-making, and process automation at scale. Yet most organizations lack the frameworks to quantify impact or the governance to move from isolated experiments into enterprise ...
Agentic AI Adoption Statistics for 2026 – First Page Sage
Key agentic AI adoption statistics for 2026, including rates by industry, company size, and implementation stage. Includes free PDF download.
Autonomous Operations: The Next Competitive Advantage
With Agentic ERP, businesses can prepare for growth without being limited by outdated systems or manual processes. Organizations that fail to modernize their operations may find it difficult to compete in a faster and more data-driven business environment. Without AI-driven automation, ...
Building DoorDash Assistant: An engineering overview
DoorDash details the development of an AI assistant using personalization and memory to support consumer decisions like grocery selection. It serves as a production example of agentic AI in consumer platforms.
The Rise of Agentic AI: What Businesses Need to Know
The Rise of Agentic AI: Learn how autonomous AI agents plan, reason, and act independently to boost productivity, efficiency, and business growth.
The Strongest Teams of AI Agents Will be Built Using Different Models
This HBR article argues that effective AI agent teams require diverse models, data sources, and governance approaches.
Tencent tests new AI agent Xiaowei on WeChat
Known as ‘Weixin’ in China, WeChat has rolled out the AI agent on a phased basis. Read more: Tencent tests new AI agent Xiaowei on WeChat
Researchers introduce Self-Harness, a framework that lets AI agents rewrite their own rules, boosting performance up to 60%
Not every company can or should build their own frontier AI language model. However, the harness controlling the model is something that most enterprises can and should customize for their specific purposes. Of course, this is easier said than done. Agent harnesses are still largely tuned through manual, ad hoc debugging — a process that relies heavily on intuition rather than systematic feedback loops, making it difficult to keep pace with rapidly evolving LLMs. To solve this challenge, researchers at the Shanghai Artificial Intelligence Laboratory have introduced “Self-Harness,” a new paradigm in which an LLM-based agent systematically improves its own operating rules. By examining its own execution traces to apply edits, the system trades manual guesswork for empirical evidence. Self-improving harnesses can enable development teams to deploy robust custom agents that continually adapt their own execution protocols to overcome model-specific weaknesses. The challenge of harness engineering An LLM-based agent's performance is not determined solely by its underlying base model, but also by its harness: the surrounding system that provides context and enables the model to interact with the environment. A harness includes components like system prompts, tools, memory, verification rules, runtime policies, orchestration logic, and failure-recovery procedures. This layer is crucial because many common agent failures stem from the harness rather than the model. For example, an agent may report success without checking the model’s response (e.g., running the code to see if it passes the tests), or it might retry a failed action repeatedly. The harness is also responsible for preventing context rot or overload when the agent’s interaction history grows very large. Examples of popular harnesses include SWE-agent, Claude Code, Codex, and OpenHands. Harness engineering remains a significant challenge, but the bottleneck isn't necessarily that humans are too slow or incapable. In fact, Hangfan Zhang, lead author of the Self-Harness paper, told VentureBeat that "in many cases, an experienced engineer with deep domain knowledge can still propose better changes than an LLM can today." Instead, the true bottleneck of manual engineering is that it relies heavily on ad hoc debugging rather than a verifiable, empirical feedback loop. "The deeper issue is that the current harness-engineering paradigm often lacks a systematic feedback loop," Zhang explained. "Many edits are made based on intuition, a few observed failures, or ad hoc debugging." With new models being released at a rapid pace, depending on human intuition to manually tune model-specific harnesses becomes increasingly costly and untenable. While some approaches use stronger models to improve the harnesses of weaker target agents, this dependence on external guidance has its own challenges, as these models may be costly, unavailable for frontier models, or mismatched to the target model's failure modes. How Self-Harness works The Self-Harness paradigm enables an LLM-based agent to improve its own harness without relying on human engineers or stronger external models. This continuous self-evolution is driven by a three-stage iterative loop that turns behavioral evidence into harness updates: Weakness mining: Starting from an initial harness, the agent runs a set of tasks, producing execution traces with verifiable outcomes. The agent categorizes failed traces and tries to detect model-specific failure patterns. Harness proposal: Based on these failure patterns, the agent uses a “proposer” role to generate a set of diverse yet minimal harness modifications, each tied to a specific failure mechanism to avoid overly general corrections. Proposal validation: The system evaluates candidate modifications through regression tests. An edit is promoted only if it improves performance without causing measurable degradation on held-out tasks. If multiple candidate modifications pass the regression tests, they are merged into the next version of the harness, which then serves as the starting point for the next iteration. To visualize why an enterprise would need this, imagine an automated issue-fixing agent that reads internal documentation, writes patches, and opens pull requests. If the company updates its documentation style, the agent might suddenly fail, pulling the wrong context or writing bad patches. On the surface, the agent simply looks broken. But Self-Harness turns this ambiguous failure into a solvable problem. "The failure traces expose where the agent is misusing the new documentation format; the proposer can generate a targeted harness edit... and the evaluator can decide whether that edit improves the failing cases without regressing other cases," Zhang said. Self-Harness in action The researchers evaluated Self-Harness on Terminal-Bench-2.0, a benchmark that tests general tool-based execution, including artifact management, command use, verification behavior, and recovery from execution errors. They applied Self-Harness with MiniMax M2.5, Qwen3.5-35B-A3B, and GLM-5. To isolate the impact of the self-evolving harness, they started with a minimal harness built upon the DeepAgent SDK, containing only the benchmark-facing system prompt, and the default filesystem and shell tools. The model backend, tool set, benchmark environment, and evaluator were kept unchanged while only the harness was allowed to vary. The quantitative results show that agents improved their performance through automated harness edits. On held-out tasks, performance jumped significantly across the board, ranging from 33 to 60 percent relative improvements for different models. Importantly, an explicit acceptance rule promotes only those edits that improve performance without introducing unacceptable regressions. What makes Self-Harness powerful for enterprise applications is that it doesn’t simply make the prompt longer or add generic instructions. Instead, it introduces targeted changes that reflect the recurring problems each model encounters during execution. For example, under the baseline harness, MiniMax M2.5 would get stuck endlessly exploring dataset configurations until the execution environment timed out, failing to produce any deliverables. Through Self-Harness, the system identified this specific flaw and wrote a "loop breaker" into its runtime policy, forcing the agent to stop and redirect its approach after 50 tool calls. It also added a rule to create an initial version of required artifacts as early as possible. On the other hand, Qwen-3.5 had a habit of hitting a file overwrite error and then blindly retrying the same command repeatedly, eventually deleting necessary files out of confusion before stopping. The self-harness fixed this by introducing a strict command-retry discipline (forbidding exact duplicate commands) and a mechanism that forced the agent to immediately recreate any missing artifacts if a file error occurred. GLM-5 struggled to preserve environment changes across different commands, and would often waste time on massive downloads or finalize tasks even when sanity checks were failing. Its self-generated harness introduced rules instructing the agent to persist PATH variables across shell sessions, limit external compute, and repair any failed sanity checks before concluding its run. The hidden costs of automated harnesses While Self-Harness automates the tedious work of tracking down idiosyncratic model failures, decision-makers must be realistic about the trade-offs. Replacing human engineering with automated trial-and-error requires significant computational overhead. "Self-Harness replaces part of the human engineering burden with repeated proposal generation, parallel candidate evaluation, and regression testing," Zhang said. "That can mean more API tokens, more latency during optimization, and more infrastructure for running evaluation tasks." Also, this system relies on the accuracy of its evaluation pipeline. During their experiments on Terminal-Bench-2.0, the researchers relied on strict, deterministic verifiers to ensure the agent's edits were actually helpful. Without this rigorous ground truth, an automated system risks promoting bad updates. "[The] evaluation system is not an optional component; it is what lets us trade human intuition for empirical evidence," Zhang said. This reliance on strict verifiers also dictates where Self-Harness should be deployed. "The best deployment targets today are environments where failures can be measured and where trial-and-error is relatively safe," Zhang said, pointing to coding, internal workflow automation, and DevOps data pipelines as ideal use cases. Conversely, enterprises should avoid fully automating harnesses in high-stakes or subjective fields. "The clearest red flags are domains where evaluation is subjective, delayed, non-deterministic, or costly to get wrong, such as medical decision-making, safety-critical infrastructure, or legal decisions." From prompt tweakers to feedback architects The introduction of self-improving agents does not mean coding or enterprise workflows will suddenly become human-free. The quality of collaboration between the human engineer and the AI is still paramount and difficult to capture with automated benchmarks. Instead, the engineering profession is moving up the abstraction layer. "The role of enterprise engineers will shift from manually patching individual prompts or tool calls toward designing the feedback systems that make agent improvement possible," Zhang predicted. Moving forward, "the engineer becomes less of a prompt tweaker and more of a feedback architect." As foundational models grow more capable, they will naturally absorb many capabilities that currently require manual harness engineering. "But once that happens, the harness will not disappear; its scope will move outward to connect the model to richer external environments," Zhang said. "Until that boundary moves beyond what humans can evaluate, humans will remain critical providers of feedback."
Nvidia gets all agentic about supercomputing for scientific research
Tireless AI agents could help scientists do research humans alone can't, says GPU giant
Enterprise AI Agent Security Gets New Architecture: AWS Continuum and Context
Enterprise AI agent security gained new infrastructure June 17 as AWS launched Continuum and Context — two services addressing the documented failure of AI agents in production, triggered in part by the episode in which Claude Mythos forced the U.S. government to issue its first export control
AI Development Shifts: Loop Engineering Replaces Manual Prompting, Experts Advise Embracing Autonomous Systems
The industry is moving toward loop-based AI development where agents autonomously refine prompts, sparking debates over terminology like 'vibe coding' versus 'agentic engineering.'
AWS Continuum offers devs help with securing code | CSO Online
Continuum is a new service intended to hep developers and security teams secure their own code and that of others too, with a goal of automating remediation.
Waymo hits the brakes after robotaxis keep missing the signs for freeway construction zones
Nearly 4,000 vehicles recalled for driving past closure warnings and between cones marking shut lanes.
Tensordyne makes a big bet on log math to beat Nvidia
Who needs compute-hungry multiplications when you can just add logarithms.
GOOGL Stock Dips Premarket After Breakout Week: Analyst Says Google Developing Next-Gen AI Chip With MediaTek
Google’s “Triggerfish” will feature a substantial memory upgrade, support next-generation HBM4E memory, and allow more active workloads to remain on-chip, according to Ming-Chi Kuo.
Cities must set the terms for the AI infrastructure boom – POLITICO
With the scale of investment flowing ... clean energy rather than lock in new fossil fuel demand. ... These are practical, city-led solutions to a global challenge. Investors and tech companies need clarity so they can expand responsibly and invest in our cities. Residents need confidence that their voices will be heard and that they will see the benefits of these developments. Strong city leadership provides both. The companies building the infrastructure of the AI age now face ...
Data centers get a faster path to power, but there's a cost - TheStreet
Speeding up grid connections solves one problem for AI infrastructure. It may create a new one for everyone else's electric bill.
AI boom enters build phase, lifts infrastructure, pressures software | Advisor.ca
CIBC Global Asset Management’s Mickey Ganguly says AI’s trillion-dollar infrastructure spend in 2027 will drive new winners beyond big tech.
3 US AI Infrastructure Stocks With Backlog Growth And Funding Risk - Simply Wall St News
AI infrastructure is where artificial intelligence meets concrete, cables, and cooling, and that physical backbone is attracting serious attention as investors reassess risk around inflation, rates, and growth across regions. While bond markets react to shifting central bank signals and consumer ...
China’s MLCC suppliers eye Hong Kong capital as AI reshapes electronics supply chains | South China Morning Post
Small, found in almost every electronic device and critical to AI servers, now the ‘rice of electronics’ is driving two Hong Kong IPOs.
China's push for green power use in AI projects faces hurdles, experts say | Reuters
China's drive to ramp up renewable power for its fast-expanding AI data centre sector is running into hurdles, as industry experts warn that forecasting peak demand remains difficult and grid operators are wary of taking on added risk.
China’s AI power crisis: Green energy goals hit major hurdles
Ensuring reliable electricity for ... infrastructure and power supply networks. China’s AI power crisis: Green energy goals hit major hurdles · A key part of that effort is an ambitious plan to channel more green electricity directly into the rapidly growing data centre industry. Authorities aim for renewables to supply four-fifths of the sector's total power consumption by 2030, a ...
US Acts to Speed Up Power Grid Hook-Ups for AI Data Centers
US regulators have taken their biggest step yet to speed the connection of data centers to the country's grids while simultaneously attempting to slow
Chevron moves into power production with Microsoft AI deal
Company signs 20-year agreement to develop data centre in heart of US oil country that could include gas-fired plant
Masayoshi Son Dismisses Musk’s Idea for Orbital Data Centers
SoftBank Group Corp. founder Masayoshi Son said there’s little merit to building data centers in space, as championed by Elon Musk, predicting that the AI race will be clinched by compute on Earth.
AI hit the memory wall — now it needs a new context tier
Presented by Solidigm As inference workloads evolve from discrete question-and-answer exchanges into persistent, multi-step agentic systems, GPU availability is no longer the most critical AI bottleneck. Instead, the bottleneck has migrated from compute to context, says Jeff Harthorn, AI applied research lead at Solidigm. "Why context management has become a primary bottleneck, more than GPU availability or compute efficiency, is the question of 2026," says Harthorn. "GPUs have gotten dramatically cheaper per FLOP. Model architectures and inference serving engines have all gotten much more efficient. But the thing that's grown faster than both of those is context. The persistent state that has to live between sessions has grown even faster than context itself." It's happening as context windows grow dramatically, making individual inputs far larger than before. Agentic AI systems chain dozens or hundreds of model calls together, each generating state that must be tracked, and enterprises are requiring that inference state persist across sessions for audit, governance, and reuse. These trends compound each other, pushing context volumes beyond what any existing memory tier was designed to handle. "Those three things are all happening at the same time, all of which are pushing context data and context memory into the stratosphere much more quickly than we're used to seeing," adds Ace Stryker, director of AI and ecosystem marketing at Solidigm. The solution is a dedicated context tier emerging between GPU memory and bulk network storage: a layer of high-performance, high-density flash designed specifically to hold and serve Key-value (KV) cache, the inference data that allows models to retain and reuse context, and retrieval data at inference speed. Nvidia has formalized this architecture under the term CMX. Storage companies including Solidigm are building SSD products optimized for this workload. "Storage has not been the first thing folks have thought about when they've been planning their enterprise infrastructure buildout," Stryker says. "In a lot of ways, it was a relatively small cost compared to compute, and it was a commodity. You just shopped around for the lowest dollar per gigabyte and called it good. But now, if your storage is not up to snuff, your ROI suffers, and it directly impacts your bottom line.” Why AI inference requires a different storage architecture than training The storage architecture that AI systems rely on today was largely inherited from training workflows. Training is sequential and write-dominated, with data moving in large blocks to and from bulk object storage. The tier structure, with high-bandwidth memory on the GPU, fast NVMe in the server, and bulk storage over the network, serves that use case reasonably well. However, inference is a different animal. Its I/O signature is fine-grained, latency-sensitive, and increasingly stateful. KV cache data and retrieval data each have distinct access patterns, but both need to be served quickly and reused across interactions. Neither fits cleanly within GPU high-bandwidth memory, which is expensive and physically constrained, nor within traditional bulk storage, which was never designed for active inference workloads. "The architectural gap that's interesting to me right now isn't at the top of the stack or the bottom, it's right in the middle," Harthon says. "A lot of what sits below the GPU HBM is being asked to do things it wasn't really designed for, which is where the most interesting systems work today is happening." One of the most visible symptoms of this gap is recomputation. In inference, the pre-fill stage processes all of the context relevant to a given session before token generation can begin. When KV cache state isn't available in a fast, accessible tier, the system recomputes it — burning GPU cycles that produce no new value. "A meaningful share of GPU cycles end up going to re-pre-filling," Harthon explains. "During all of that calculated context, that's potentially compute that's being spent reproducing state, rather than doing new work. When you start looking at the problem that way, GPU utilization starts looking like it's partly a storage problem." This reframing is driving renewed interest in a metric borrowed from networking: goodput, or useful tokens per dollar, rather than raw tokens per dollar. The AI context memory tier and how it works The industry's response is taking structural form. A new tier is emerging between GPU memory and traditional network storage, designed specifically to hold and serve inference context, a layer distinct from drives inside GPU servers (G3) and storage servers over the network (G4), engineered to serve context data back to accelerators as rapidly as possible. "If you're building a data center starting in the second half of this year, or the beginning of next year, you can't think about storage only living in two places," Stryker says. "Storage has to live in at least three places to handle the context memory tier, and that's likely to be a permanent fixture in how the infrastructure gets built going forward." It's analogous to the emergence of object storage as a category, which didn't exist until enough workloads needed it. And once it did, it developed its own primitives, SLAs, cost models, and an ecosystem of vendors. "The context tier looks like it might be on a similar arc," Harthorn says. "That volumetric pressure is causing the category to form, rather than any one vendor's road map." For infrastructure leaders, this means actively planning for the new tier rather than treating it as optional. Deploying additional NAND at this layer reduces dependency on DRAM, which is orders of magnitude more expensive per gigabyte and constrained in both availability and thermal headroom. "In terms of your investment effectiveness, you're laying out less cash to do it if you rely on the SSD layer in the way that Nvidia is now recommending and prescribing for a lot of use cases," Stryker adds. What flash needs to deliver to support AI inference Participating meaningfully in the inference stack places new demands on SSD technology. Tail latency, the worst-case performance of a drive, must be predictable, not just fast on average. An orchestration system that allocates GPU resources based on expected storage response times cannot tolerate unexpected multi-second delays. Consistent, observable performance matters more here than peak throughput. Beyond latency, density becomes a critical concern, especially at hyperscale. In data centers where power, not cost, is the binding constraint, watts per petabyte becomes the operative metric. Floating gate NAND, the manufacturing approach at the core of Solidigm's products, is suited to that calculation. Network integration via NVMe over Fabrics, RDMA, and eventual CXL support is also essential, given the tight latency budgets of active inference pipelines. "The drives have to have reliable performance characteristics, beyond the throughput side and being able to transfer as much data as possible as fast as possible, the way that training needed," Harthon says. "Now it's about being able to do it very consistently, in a way that's very observable to the people operating and orchestrating these systems." How enterprise AI leaders should plan for the context tier The standards, software primitives, and best practices being established now will define how AI inference infrastructure operates for years to come. Solidigm is engaged in that process through standards bodies, partner lab collaborations, and published research, which is critical precisely because the category is still forming. "The interesting question for the next couple of years isn't whether AI infrastructure needs more compute," Harthorn says. "It's whether it can use what it has more efficiently. A lot of that answer runs through this tier that is being built today." Sponsored articles are content produced by a company that is either paying for the post or has a business relationship with VentureBeat, and they’re always clearly marked. For more information, contact sales@venturebeat.com.
AI Data Center Market Projected to Reach USD 810.6 Billion by 2033 as Enterprises Accelerate Investments in AI Infrastructure | The AI Journal
Growing Demand for Generative AI, High-Performance Computing, and Cloud-Based AI Workloads Fuels Rapid Expansion of Next-Generation Data Centers
AI’s biggest bottleneck is no longer chips — it is electricity
As AI models grow larger and more ... biggest challenge facing technology companies is securing enough reliable power to run sprawling data centres. Training frontier AI models and serving billions of user queries require enormous amounts of continuous electricity, pushing power grids to their limits and forcing companies to rethink where the next generation of computing infrastructure should be built. That growing energy crunch has ...
The second half of the computing power battle: Intel CEO Pat Gelsinger reveals how AI is reshaping the global semiconductor supply chain｜GPU, semiconductor - ChainCatcher
Intel CEO Pat Gelsinger believes that the growth in AI demand has surpassed GPUs, involving bottlenecks in power, memory, packaging, and more. Intel's revival requires systematic repair through the balance sheet, engineering culture, and customer trust. CPUs are regaining importance in Agentic ...
AI demolishes traditional tech: how NPUs and AI RAN are rewriting European infrastructure
AI is no longer a localized software novelty. It is now aggressively wiping out traditional hardware infrastructure across Europe. According to new market intelligence reports from CONTEXT World, there has been an unprecedented displacement of legacy systems.
How Many Barrels of Oil Do AI Data Centers Consume on a Daily Basis? | OilPrice.com
AI companies are now scrambling for the same thing oil companies have fought wars over: secure access to energy. JLL estimates global data-center capacity will nearly double by 2030, requiring almost 100 gigawatts of new supply and as much as $3 trillion in combined infrastructure and GPU spending. The International Energy Agency (IEA) projects global data-center electricity demand ...
Microsoft (MSFT) And Chevron Build A Huge Power Backing For AI Data Centers
Chevron and Microsoft announced a large co-located natural gas power plant and data center project in the US, focused on supporting AI infrastructure. The project is designed to secure dedicated energy capacity for Microsoft's growing AI data center workloads.
Chevron to Fuel Massive Microsoft Data Center in Texas With Natural Gas in 20-Year AI Power Deal | IBTimes
Chevron has announced a landmark agreement with technology titan Microsoft to supply natural gas-fired electricity to a massive artificial intelligence data center campus in West Texas.
NVIDIA Expands AI-HPC in Europe with 35 New Systems
NVIDIA is boosting Europe's AI landscape with 35 new systems aimed at scientific advancement, including upgraded supercomputers like MareNostrum 5 and Blue Swan.
India Leads Asia's Data Center Boom Amid AI Demand
India is poised to become one of Asia's fastest-growing data center markets due to rising AI demand and strategic location.
MGX could purchase APAC data center operator DayOne – report
Abu Dhabi investment firm MGX is reportedly looking to purchase Singaporean data center operator DayOne. Citing three unnamed sources, Reuters said MGX was working with an investment bank in preparation for a potential transaction. A DayOne data center in Finland – DayOne Spun out of Chinese firm GDS in 2025, APAC-focused DayOne has been aiming […]
Nvidia says its AI data center design runs hotter to use a lot less water
Nvidia's new data center cooling designs prioritize water conservation by allowing hardware to operate at higher temperatures.
China Prepares a National Industry to Bring AI Data Centers to Space | Cloud News
China wants the next wave of data centers not to be limited to land-based infrastructure. Beijing has approved the creation of the Space Computing Industry
Utah Data Center Brute Forced Through to Approval
Local commissioners in Box Elder County approved a controversial data center project despite significant public opposition.
Data centers are a proxy
New polling shows that data centers have become a physical symbol of wider AI anxiety, with nearly half of respondents supporting a temporary construction ban.
Sarenet inaugures its new data center at the Derio communication hub | Cloud News
Moreover, the company highlights the importance of data sovereignty in AI development. It ensures that client information remains under their control, hosted on European infrastructure and protected by EU regulations, safeguarding privacy, security, and regulatory compliance.
The new database world according to Google: Inexact queries and AI in everything
'Humans are not going to be using data platforms in the next three to five years,' product exec tells us
Meet Fugu: The AI Model That Doesn't Answer Your Question, It Hires a Team
Instead of betting on one giant model, Sakana Fugu dynamically assembles a team of specialist models behind a single API - and says its Ultra tier now competes with frontier systems on coding, science, and reasoning.
AI Exchange
In this monthly series, running alongside our existing Tech Exchange dialogues, FT journalists talk to the scientists, developers and business leaders exploring ever more applications for artificial intelligence in every aspect of our lives
What is Sakana AI's "Fugu"? The True Nature of "Orchestration AI" Revealed Through Comparison with Perplexity Computer and Claude Mythos｜佐藤源彦@MBBS
Hello everyone. This is AI Co-creation Innovation Wonder Sato."In the future of AI, the company that builds the largest model will win"—I think many of you feel this way.It is true that AI development until now has been a race to spend vast amounts of money to nurture a single, super-large model
AI models capable of devastating attacks on governments and business months away, rare Five Eyes statement warns
Signal agencies in Australia, the US, the UK, New Zealand and Canada sound alarm after Trump blocks foreign nationals from Anthropic’s Fable AI model Powerful AI models capable of devastating new cyber attacks on governments and businesses are mere months away, intelligence agencies for the Five Eyes have warned in a rare joint statement, urging leaders to “act now”. The surprising public intervention by signals agencies for Australia, the US, the UK, New Zealand and Canada comes after the Trump administration earlier this month decided to block “foreign nationals” from using a much-hyped AI model built by tech company Anthropic, called Fable. Continue reading...
AI-powered threats may succeed ‘within months’, Five Eyes warns
Western governments and corporates cautioned that their lead in AI might not last long
Agentic AI in Cybersecurity: How autonomous AI is transforming the protection of critical infrastructure
Agentic AI is transforming cybersecurity through autonomous systems capable of detecting, deciding, and responding to threats in real time.
What The New Executive Order On AI And Cybersecurity Means For Your Business - New Technology - United States
The EO responds to concerns surrounding two recently announced next-generation frontier AI models — Anthropic’s Mythos and OpenAI’s 5.5 Cyber — which are currently in limited testing and have not been released...
Council Post: How AI And Quantum Computing Are Rewriting Cyber Risk
The same weaknesses leave organizations exposed to both AI-enabled attacks and delayed cryptographic migration.
Sniff out stale AI override advice with this open source CLI
Package dependencies can create vulnerabilities that are fiendishly hard to find and stamp out
Five Eyes security agencies call for urgent action to manage AI risk
Cybersecurity agencies from the US, UK, Canada, Australia, and New Zealand warn that AI will rapidly transform cyber capabilities and urge organizations to prioritize resilience.
OpenAI: Yoo-hoo, look over here, we do that security stuff too!
A plethora of pwn-prevention, including a 'Patch The Planet' pledge
MosaicLeaks: Can your research agent keep a secret?
ServiceNow introduces MosaicLeaks, a benchmark for privacy leakage in deep research agents. It highlights risks like intent and information leakage in enterprise LLM use cases.
Read this before you vibe-code another app
A look at the security risks associated with 'vibe-coding' and the rapid development of AI-assisted applications.
OpenAI launches new initiative to help find and patch open-source bugs
OpenAI has introduced a program aimed at identifying and fixing security vulnerabilities within open-source software projects.
Adoption, Deployment & Impact
Drowning in AI: Companies are launching hundreds of projects, and that’s a problem
At Fortune Brainstorm Tech, executives from Carvana, GXO, and other firms shared lessons for companies about rapid AI adoption.
Powering Up: How the Energy Sector is Scaling AI and Cyber Resilience in 2026
However, unlocking full value remains a challenge. Legacy infrastructure, data readiness gaps and governance complexity continue to slow progress, reinforcing the importance of strong data foundations and responsible AI frameworks. Anish De, Global Head – Energy, Natural Resources and Chemicals ...
Mortgage AI & Automation 2026: Speed Is Losing to Proof and Auditability | Windows Forum
Pennymac’s latest Policy Pulse and Mortgage News Daily’s June 22, 2026 roundup describe a mortgage market where lenders are being squeezed simultaneously by policy volatility, higher-for-longer rates, AI compliance anxiety, verification bottlenecks, and a fix-and-flip segment that looks...
New AI era defined by agents, rising costs and maturity gaps
Enterprises face higher AI bills and governance gaps as only 17 per cent have reached high maturity, Gartner says.
Are your legacy systems hindering AI implementation? | The Drum
Let’s be direct: most enterprise technology was not built for AI. It was built for stability, predictability and scale in a world of batch-oriented workloads and siloed data. That world still exists inside most organizations – and it’s quietly throttling every AI initiative you are trying ...
IT teams are bullish on AI tools, but they’re worried security practices can’t keep pace | IT Pro
Executives and IT teams are at odds over the risks associated with AI adoption
Why Enterprises Struggle to Scale AI to Production - The European Business Review
Explore the challenges enterprises face in moving AI from pilots to production and achieving scalable business value.
AI is transforming enterprise data risk. Here’s how security leaders are responding. | Cybersecurity Dive
New research from 1,700 security leaders reveals 3 imperatives for securing AI adoption.
Micro AGI's 'Shift' Initiative: Free Services for In-Home Data Sparks Privacy Concerns
Micro AGI is offering free home services in exchange for data collection to train AI robots, raising concerns among privacy advocates regarding data misuse.
Eli Lilly deploys weight-loss cash on ‘App Store’ for scientists
Mounjaro maker is collaborating with small biotechs on AI as a tool for drug discovery
Nvidia’s Kimberly Powell: We are reinventing the doctor experience
The chipmaker’s head of healthcare argues AI can ease many of the sector’s ills, including reducing medics’ workload and tackling the shortage of trained staff
1 in 3 Americans use chatbots for health advice. These 6 patients explain why. - The Washington Post
They say cost, speed and other considerations drove them to AI . They are also learning what to believe and what to take with caution.
Insilico Medicine, SK Biopharmaceuticals strike $2.5B AI drug discovery deal targeting neuroimmune therapies
“We want to be the SpaceX of the pharmaceutical industry,” Alex Zhavoronkov, co-CEO of Insilico Medicine, tells Fortune.
LATAM Airlines Leverages AI for Major Digital Overhaul, Boosting Efficiency and Customer Experience
LATAM Airlines launched an AI-driven digital transformation on June 20, 2026, to improve routing, maintenance, and customer service operations.
Brands using AI-generated influencers to promote products on social media
Companies are increasingly utilizing AI-generated influencers to market products, raising questions about transparency and authenticity.
60% surveyed professionals say AI now central to HR operations: Report, ETHRWorld
Artificial Intelligence In HR: The 'AI As The New HR Priority -- Efficiency, Cost and Workforce Impact' report is based on a survey among 1,811 HR professional across industries conducted between May 7-31.
London’s Isometric lands €34 million to scale certification platform for industrial markets
Isometric has closed a €34 million ($40 million) Series A to expand its AI certification platform across the €305 billion ($350 billion) industrial certification market and accelerate its AI-enabled services expansion. The round is led by AVP. All of Isometric’s existing institutional investors, including Lowercarbon Capital and Plural, joined the round, with personal investment from Kleiner […]
Where AI Actually Lands in the M&A Strategy Phase — Dr. Karl Michael Popp
Neues Buch. Grundlage für die Digitalisierung des M&A-Prozesses. Alle Aufgaben mit Automatisierbarkeit, alle Fragen.
Why agentic enterprises need to become learning systems
Presented by Splunk Every day, organizations learn things their AI systems never get to use. A security analyst corrects an AI-generated investigation. A network engineer identifies the root cause of a recurring outage. An observability team discovers that a pattern of latency, logs and infrastructure changes predicts service degradation. A customer operations team learns which signals indicate an escalation is likely. Each moment contains valuable organizational knowledge. But in most enterprises, that knowledge disappears into tickets, dashboards, chat threads, post-incident reviews and the minds of individual experts. It may help solve the immediate problem, but it rarely becomes part of a reusable system that improves future AI-driven decisions. That is the next challenge for the agentic enterprise. The future will not be defined simply by who has the most capable model or the most autonomous agents. Many organizations will have access to similar frontier models. Many will deploy agents across security, IT, engineering, customer service, and business operations. The real differentiator will be whether those agents can learn from the organization around them. Not by constantly retraining the underlying model, but by capturing operational experience, converting it into institutional knowledge and making that knowledge available to future agents, workflows, and decisions. The agentic enterprise is not just an enterprise that uses AI. It is an enterprise that learns through AI. Agentic enterprises allow AI systems to learn from them The AI conversation has been dominated by model capability: larger context windows, better reasoning, faster inference, stronger tool use, and more sophisticated agentic behavior. Those advances matter. But in the enterprise, a model is only one part of the system. A model does not automatically know how a specific organization operates. It does not inherently know which remediation step solved last month’s outage, which analyst correction improved a threat investigation, which network signal preceded a service disruption, or which internal policy should override an otherwise plausible recommendation. That knowledge belongs to the enterprise. For agentic systems to improve, organizations need a way to capture that knowledge and make it reusable. In many cases, that does not require changing the model itself. It requires changing the ecosystem around the model: the knowledge base, retrieval layer, prompts, policies, guardrails, routing logic and workflows that shape how agents behave. The model may remain the same. The learning system around it becomes smarter. Feedback loops turn every outcome into a teachable moment for agents Every agentic workflow creates signals. An agent receives a request. It retrieves context, reasonsthrough possible actions, calls tools, and generates answers. A human accepts, rejects, or modifies that answer. Downstream systems reveal whether the action worked. That entire chain is valuable. AI observability gives organizations visibility into what happened: the prompt, response, reasoning path, tool calls, data sources, intermediate steps, failure modes and outcomes. Without that visibility, organizations cannot understand why an agent behaved the way it did, let alone improve it. But observability alone is not enough. The larger opportunity is to turn observed behavior into institutional knowledge. A trace should not only help a developer and operators debug an agent. It should help the enterprise understand what the agent learned, what the human corrected, what outcome followed, and what should change before the next similar event. That is the shift from monitoring AI to teaching AI. In the agentic enterprise, feedback loops connect action to outcome, outcome to knowledge and knowledge back to future action. A learning system in practice across security, observability and the network Consider a service experiencing intermittent degradation. An observability agent detects unusual latency and error rates. A network agent identifies packet loss across a specific path. A security agent notices that the same time window includes suspicious authentication behavior and unusual traffic from a previously unseen source. Individually, each agent has only a partial view. Together, they create a richer operational picture. The first time this incident occurs, human experts may need to intervene. A network engineer confirms that packet loss was caused by a misconfigured routing change. A security analyst determines that the suspicious traffic was not an attack, but a side effect of a misrouted internal service. An SRE connects the network event to the application degradation. That resolution contains knowledge the organization should not have to relearn. A mature agentic learning system would capture the traces, human corrections, topology context, security findings, observability signals and final remediation steps. It would preserve the relationship between those signals: latency pattern, network path, identity behavior, routing change and remediation. The next time a similar pattern appears, agents would not start from zero. They could retrieve the prior case, compare current conditions, recommend the proven diagnostic path and escalate with better context. The underlying frontier model did not need to be retrained. The enterprise learned. The architecture of the learning agentic enterprise A learning-oriented agentic enterprise needs more than a model or chatbot. It needs an architecture that can capture experience, turn it into usable knowledge, connect that knowledge to operational context, and govern how it changes future agent behavior. Memory preserves what happened: what the agent saw, what it did, where humans intervened, and what outcomes followed. Knowledge bases turn that experience into reusable guidance, including playbooks, examples, policies, procedures, and evidence. A data fabric connects the operational environment. The signals agents need live across logs, metrics, traces, tickets, identity systems, security tools, network telemetry, collaboration platforms, and business applications. A data fabric makes those signals discoverable, correlated, governed, and usable in context. AI observability explains how agents behave by capturing prompts, tool calls, intermediate steps, responses, feedback, and outcomes. That visibility helps organizations understand where agents succeed, where they fail, and what should improve. The control plane governs how learning becomes change: what knowledge is promoted, which prompts or policies are updated, which agents can use new information, what approvals are required, and how changes are audited. Together, these capabilities allow AI systems to improve over time in a controlled, trustworthy way that allows the enterprise to learn from its own operations. The organizations that learn fastest will win The next era of AI will not be won by models alone. It will be won by organizations that can capture what they learn from every workflow, expert correction, incident, investigation, and outcome. The most advanced agentic enterprises will not simply deploy more agents. They will build systems that allow every agent to benefit from the collective knowledge of the organization. That means connecting operational data through a data fabric. It means observing agent behavior deeply enough to understand it. It means preserving experience in memory and institutionalizing it in knowledge bases. It means using a control plane to govern how learning changes agent behavior. The future of AI is not a single autonomous agent acting alone. It is an ecosystem of agents, humans, data and controls that learns over time. The organizations that build that ecosystem will create AI systems that get better with every interaction. Not because the model is constantly changing, but because the enterprise itself is becoming more intelligent. Learn more about how Cisco Data Fabric powered by the Splunk Platform is accelerating agentic operations. Hao Yang is Vice President AI at Splunk, a Cisco Company. Sponsored articles are content produced by a company that is either paying for the post or has a business relationship with VentureBeat, and they’re always clearly marked. For more information, contact sales@venturebeat.com.
The Real Test Of AI Is Not Productivity. It’s Organizational Capacity.
AI is rapidly advancing, becoming cheaper and more capable, prompting a shift from model-specific strategies to model-agnostic approaches. True value lies in capacity.
Retail Pulse Report: Why AI ROI is Hard to Find
Personal benefit does not (yet) translate into retail corporate benefit – except in consumer applications. But that still has a huge caveat.
Enterprise AI adoption is at record highs. Enterprise ROI is largely unproven. Is tokenmaxxing the gap between them? — TFN
Enterprise AI spend hit $11.6M in 2026, but 56% of CEOs report no revenue or cost benefit. Investors now call “tokenmaxxing” the practice of optimising token counts instead of real outcomes.
Dataiku BrandVoice: The AI Performance Reckoning Has Arrived For CIOs. Here's The Formula Needed To Thrive
As AI becomes a board-level priority, CIOs face growing pressure to prove ROI, manage governance, and orchestrate enterprise-wide AI adoption to drive measurable business value.
CFOs Are Coming For The Enterprise AI Budget
Enterprise AI vendors know the next sale will be won not only on model quality and capability, but also on control and cost.
Council Post: AI Didn't Change The Rule. It Made It More Important.
To get real value from AI, focus on understanding the business problems that need to be solved before you reach for the tool.
Geopolitics, Policy & Governance
China Tightens Rare-Earth Grip on U.S. Firms, Threatening Trade Clash
The move targets two U.S. manufacturers at the center of the Trump administration’s effort to rebuild the domestic supply chain for critical magnets.
TechCrunch Mobility: A new robotaxi scorecard shows China’s dominance
A new industry scorecard highlights China's leading position in the global robotaxi market.
India Urged to Build Indigenous AI Models to Avoid Dependence
A new Bernstein report warns India must build its own foundational AI models to avoid falling behind the US and China, citing geopolitical risks.
South Korea launches phase two of Physical AI Alliance
South Korea's science ministry has launched the second phase of its Physical AI Alliance, reshaping the public-private body into an execution-focused platform.
Community bankers ask US officials to provide more AI regulatory clarity
Community bankers are seeking regulatory clarity on AI to combat fraud, such as synthetic identities and counterfeit checks, according to a Federal Reserve Bank of Kansas City paper.
Top carmakers warn EU tech sovereignty drive will raise costs
Brussels’ proposals to cut reliance on US Big Tech spark concerns among European carmakers
AI & Tech Brief: White House unveils quantum executive order - The Washington Post
Lawmakers on the House Energy and Commerce Committee have come to a bipartisan agreement on children’s online safety legislation. The deal shows momentum in Congress to strike an outline on an AI deal before July 4.
New York City House primary emerges as key battleground in ‘AI civil war’
AI-focused Super Pacs are spending heavily in the midterms, and half has gone to a single Manhattan congressional race The artificial intelligence industry is spending heavily in the 2026 midterms, hoping to secure influence over the technology’s first generation of legislation – and New York City’s primary has emerged as the key battleground. AI-focused Super Pacs have raised over $100m this cycle, of which $49m has been spent so far, in dozens of congressional races across the country. Half of all spending has converged on a single Manhattan race: Tuesday’s Democratic primary in the district of NY-12. Will Craft and Andrew Witherspoon contributed reporting Continue reading...
Three things to watch amid Anthropic’s latest feud with the government
This story originally appeared in The Algorithm, our weekly newsletter on AI. To get stories like this in your inbox first, sign up here. For those of you enjoying your summer unaware of Anthropic’s latest feud with the US government, here’s a recap: In April the company said it had built an AI model called Mythos…
Japan's AI copyright bargain: not Brussels, not Washington
Japan is trying to keep one of the world’s most permissive AI copyright rules intact while making opaque AI use harder to defend. Its 2026 strategy adds pressure through transparency rules and creator compensation talks.
EU AI Act Chatbot Disclosure and Deepfake Labeling: July 22 Signatory Deadline
EU AI Act Article 50 compliance becomes enforceable on August 2, 2026, requiring chatbot disclosures, deepfake labels, and AI content watermarking from every business operating in the EU — but the July 22 signatory deadline for the Code of Practice is the earlier action point that determines
The U.S. Is Losing the AI Credibility War—to Itself | Council on Foreign Relations
Recent restrictions on advanced AI models are undermining critical U.S. cybersecurity outcomes. The administration needs a coherent strategy to work with the private sector, stay ahead of adversaries, keep allies on board, and maintain trust in the U.S. AI stack.
China's platform-economy blueprint promotes coordinated growth
China's latest platform-economy blueprint sets out a three-year framework for strengthening coordination among enterprises, including measures for AI development and platform governance.
Does Vietnam’s AI Law signal a new model of AI governance for ASEAN? - MediaLaws
This article argues that, while borrowing the EU’s formal risk-based architecture, Vietnam’s AI Law reconfigures it through expansive administrative discretion, shaping a distinct regional trajectory that may influence ASEAN’s divergence from the EU’s normative governance.
A German Court Made Google Liable For What Its AI Says About You
What happens when a court holds Google liable for its AI Overview's claims? Answer engines get cautious, and ambiguous businesses stop getting named.
Proposed New Jersey Legislation Signals Expanding AI Compliance Obligations
Like many jurisdictions across the United States, the New Jersey Legislature is advancing a cluster of proposed, sector-specific bills to regulate the use of generative artificial intelligence. These bills target areas ranging from real estate advertising and election-related communications ...
The AI shift in cyber risk: why leaders must act now | National Cyber Security Centre
Nick Andersen - Acting Director, Cybersecurity and Infrastructure Security Agency ... The AI shift in cyber risk: why leaders must act now.
Unapproved AI use shows adoption is outpacing governance - Raconteur
New data indicates big variations in the extent to which AI is being used, and also in how well it’s being used
AI-Generated Ads Should Be Exempt From EU Transparency Rules, Retail Association Says
Eurocommerce, the European retail association whose members include Amazon, H&M, Inditex, and Ikea, is asking EU tech chief Henna Virkkunen to exempt
EuroCommerce seeks ad exemption from EU AI disclosure rules
EuroCommerce has urged EU tech chief Henna Virkkunen to exclude some AI-generated advertisements from new EU disclosure rules.
AI Update— Monday, June 22, 2026
SpaceX enters Week 2 of public trading above $200 with Nasdaq-100 inclusion two weeks away. Open AI quietly shipped a major ChatGPT upgrade over the weekend. And the EU AI Act enforcement deadline is now 41 days away, with most companies still scrambling to comply.
Trump: Anthropic was national security threat
President Trump revealed in an exclusive interview that he viewed Anthropic as a national security threat, though he noted that relations have since improved.
South Korea proposes stricter substantiation rules for AI advertising claims
The Korea Fair Trade Commission has proposed amendments requiring businesses to provide evidence for AI-related performance claims in advertisements.
Reuters AI News | Latest Headlines and Developments | Reuters
Explore the latest artificial intelligence news with Reuters - from AI breakthroughs and technology trends to regulation, ethics, business and global impact.
Get the full executive brief
Receive curated insights with practical implications for strategy, operations, and governance.