Tue 5 May 2026
Daily Brief — Curated and contextualised by Best Practice AI
Palantir Surges, Meta Borrows Billions, and AI Debt Recovery Falters
TL;DR Palantir reported a $1.63 billion quarter, surpassing forecasts and downplaying AI competition concerns. Meta is arranging a $13 billion financing package for a data center in El Paso, reflecting Big Tech's growing debt dependency. Davidson Kempner warns that AI disruptions threaten private debt recovery in the software sector. Remote work and AI are reshaping job markets, affecting career mobility and skill requirements.
Economics & Markets
ServiceNow Sees $30 Billion Revenue by 2030 on AI Uplift
ServiceNow Inc. projected it would generate $30 billion of subscription revenue in 2030, attributing the strong outlook to traction from its AI products.
Inside AMEX’s agentic commerce stack: How intent contracts and single-use tokens enforce AI transactions
American Express (Amex) is building a system that lets AI agents shop and pay on behalf of users — but right now it’s only within its own payment network, and still involves a black box that could hinder trust and auditability. Amex already participates in agentic commerce protocol projects, especially Google’s Agent Pay Protocol (AP2), which focuses on interoperability. Amex’s Agentic Commerce Experiences (ACE) developer kit, on the other hand, touches on something most protocols currently lack: Full transaction control in the payment layer. But it still isn't completely transparent in how it handles validation. ACE uses a closed-loop system — serving as both the card issuer and the payment network — to validate agent-led transactions. Luke Gebb, Amex's EVP and global head of innovation, told VentureBeat that the company believes this model is the missing piece in agentic commerce. “Some of what is missing so far is the perspective of a company like ours: We feel that trust and security are critical to advancing this space,” Gebb said. “This is really the first time that an issuer is coming to the table.” Amex sits in that interesting space: Unlike other financial institutions or card providers like Chase or Bank of America, Amex can route transactions through its American Express Network. Visa and Mastercard are two of the most well-known payment networks, but these companies don’t issue cards themselves and must work with a bank. The continued black box of agentic commerce The ACE kit is just one approach to addressing some of agentic commerce’s biggest problems: trust, control, accountability, validation, and security. Consumers generally don’t want rogue agents to run away with their bank accounts and start buying things. Merchants don’t want to be stuck with unpaid items. Banks don’t want to deal with an influx of chargebacks and the potential for fraud. Projects like the ACE kit aim to build trust and accountability by verifying an agent’s identity and goals. This can build the trust agentic commerce desperately needs. Amex claims it offers validation, too, although the process behind that is unclear. It is abstracting how it performs validation, even though it explains at which layer it does it. More traditional systems feature a mix of deterministic checks and a flexible, semantic evaluation that helps match intent and outcome for validation. Amex said agents built with ACE can submit user shopping carts and check them against the agent's original intent. However, they did not disclose how this works. Practitioners building to the agentic commerce ecosystem lament that, despite strides in creating a trust layer, many black boxes remain that could hinder widespread adoption. Raj Ananthanpillai, founder and CEO of identity and verification system provider Trua, told VentureBeat that payment protocols and software kits like Agentic Commerce Suite from Stripe, Google's Verifiable Intent proof chain, and the ACE developer kit "excel at handling proofs, verifiable authorizations and the mechanics of fund movement, but leave upstream human validation opaque and underdeveloped." Ananthanpillai continued: "Without a clear, high-assurance cryptographic link proving that an agent is acting under the explicit authority of a verified human owner, merchants, issuers, and networks face heightened risks of repudiation, massive chargebacks, sanctioned people conducting financial transactions, and fraud." The ACE kit The ACE developer kit solves several running issues with agentic commerce, Gebb said, and gives developers access to integrated services: Agent registration Account enablement Intent intelligence Payment credentials Cart context First, it deals with agent registration, establishing identity and trust with both the consumer and company agents. When a transaction begins, the agent acting on behalf of the customer and the merchant’s agent can verify each other’s identities and trust that they are dealing with the correct entity. Next comes account enablement, which links the user’s Amex account to their agent and grants the agent permission to act, or, in the case of agentic commerce, buy something. Intent intelligence creates what Amex calls an intent contract, where the user defines what they want the agent to do. Once the intent is defined, the ACE system generates an Intent ID and a Proof of Intent Token that definitively proves authorization in the event of a dispute. Amex handles the actual transaction part, where the user pays for the product through a single-use token. ACE establishes payment credentials used for the transaction, bound to intent and constraints. “Once the agent has found the item that the customer has asked for, like red shoes, they'll make a call for the payment credentials, which is a token that has the boundaries that the card member has provided,” Gebb said. “So, for instance, if they said they only wanted to spend $500, that token won't allow for a purchase of $600 because it has controls built in.” The last piece is cart context and validation, which Gebb said helps banks and brands compare a user’s cart that their agent submitted to their intent. Amex’s approach shows that for agentic commerce to really soar, providers must understand what systems will allow agents to do and who is ultimately accountable if something goes wrong.
Palantir Beats Forecasts With $1.63 Billion Quarter as Sales Accelerate
On a call with analysts, the data company seeks to play down concerns about AI competition that have weighed on the stock.
Meta Taps Morgan Stanley, JPMorgan for El Paso Data Center Deal
Meta Platforms Inc. is working on a financing package for a data center in El Paso, Texas, that could total roughly $13 billion, underscoring Big Tech’s growing reliance on debt to bankroll the infrastructure behind the AI boom.
Reuters Business News | Today's International Headlines | Reuters
LegalcategoryAnthropic nears $1.5 billion AI joint venture with Wall Street firms, WSJ reports · 2:35 AM UTC · categoryWorld stocks gain, oil climbs amid new Gulf proposals · 12:59 AM UTC · categorySK Hynix shares rally 13% after US tech firms signal strong AI spending plans ·
AI + Halo = $$$
A techno-industrial boomlet
Starwood CEO on Business Strategy, AI, Data Centers
Barry Sternlicht, Chairman and CEO at Starwood Capital Group, discusses the company's business strategy and investing in AI and data centers. He speaks with Romaine Bostick from the sidelines of the Milken Institute Global Conference in Beverly Hills. (Source: Bloomberg)
☕🤖 The Pentagon Just Went All-In on AI
The Pentagon just signed classified AI deals with Open AI , Google, Microsoft, AWS and Oracle, x AI dropped Grok 4.3 with a 1M context window at fire-sale pricing, and Meta snapped up humanoid robotics startup Assured Robot Intelligence to fuel its Superintelligence Labs.
"Never Sell" Episode: Big Tech earnings, S&P and Moody’s, AI
For the “Never Sell” podcast, Scuttleblurb and I published a new episode yesterday. While I have already covered big tech’s earnings over the last few days (Meta, Alphabet, Amazon, and Microsoft), we expanded on how IT budget reallocation towards AI may not lead to any sustained long-term advantage for the enterprises.
The A.I. Industry Is Booming. When Will It Actually Make Money? | The New Yorker
As Elon Musk sues his former OpenAI partners, A.I. companies are expanding rapidly, but profits are still scarce, John Cassidy writes.
ON Semiconductor Q1 2026 slides: AI data center surge drives recovery By Investing.com
ON Semiconductor Q1 2026 slides: AI data center surge drives recovery
Geothermal startup Fervo Energy to raise up to $1.3B in IPO
Fervo Energy, a leader in geothermal power technology, is preparing for a significant initial public offering.
Corporate AI Psychosis Destroys Microsoft. Microsoft Is About to Pop the AI Bubble
Microsoft’s massive AI bets may be backfiring. Things like losses, Open AI tensions, and weak demand raise fears that the AI bubble could soon burst.
AI Threatens Private Debt Recovery in Software: Davidson Kempner
Disruptions caused by artificial intelligence are threatening private credit firms’ potential recovery rates in the software sector, according to Davidson Kempner Capital Management LP chief investment officer Tony Yoseloff.
AgentFloor: How Far Up the tool use Ladder Can Small Open-Weight Models Go?
arXiv:2605.00334v1 Announce Type: new Abstract: Production agentic systems make many model calls per user request, and most of those calls are short, structured, and routine. This raises a practical routing question that existing evaluations do not directly answer: which parts of an agent workflow truly require large frontier intelligence, and which can be handled by smaller models? We introduce AgentFloor, a deterministic 30-task benchmark organized as a six-tier capability ladder, spanning instruction following, tool use, multi-step coordination, and long-horizon planning under persistent constraints. We evaluate 16 open-weight models, from 0.27B to 32B parameters, alongside GPT-5 across 16,542 scored runs. Our results reveal a clear boundary of model necessity. Small and mid-sized open-weight models are already sufficient for much of the short-horizon, structured tool use work that dominates real agent pipelines, and in aggregate, the strongest open-weight model matches GPT-5 on our benchmark while being substantially cheaper and faster to run. The gap appears most clearly on long-horizon planning tasks that require sustained coordination and reliable constraint tracking over many steps, where frontier models still hold an advantage, though neither side reaches strong reliability. We also find that this boundary is not explained by scale alone: some failures respond to targeted interventions, but the effects are model-specific rather than universal. These findings suggest a practical design principle for agentic systems: use smaller open-weight models for the broad base of routine actions, and reserve large frontier models for the narrower class of tasks that truly demand deeper planning and control. We release the benchmark, harness, sweep configurations, and full run corpus.
Hedge funds seek an edge by using AI’s speed
Investors are using the technology to analyse documents but are holding it back from more sensitive tasks
Chegg, Penske Media seek oral argument on Google's motions to dismiss US antitrust claims
Chegg and Penske Media are seeking oral argument on Google's motion to dismiss their lawsuits, arguing their claims differ from other publisher cases in outcome-determinative ways.
Cybersecurity M&A Roundup: 33 Deals Announced in April 2026 - SecurityWeek
Significant cybersecurity M&A deals announced by Airbus, Cyera, Fortra, Palo Alto Networks, Silverfort, and Socket.
Weixin Celebrates a Decade of IP Protection with New 2025 Brand Protection Report
Weixin's 2025 Brand Protection Report highlights a shift from reactive takedowns to proactive IP protection, showing a 5.7-fold increase in enforcement value.
Becoming Immutable: How Ethereum is Made
arXiv:2506.04940v4 Announce Type: replace Abstract: Blockchain's economic value lies in enabling financial and economic transactions without relying on trusted, centralized intermediaries. In practice, however, transactions pass through a fragmented chain of intermediaries before being included on-chain. Because standard blockchain data reveal only the winning block, this process is largely unobservable. We address this limitation by constructing a novel dataset of 15,097 non-winning Ethereum blocks, that is, blocks proposed but not selected for inclusion. We show that 21% of user transactions are delayed: they appear in candidate blocks but not in the winning block, implying that fragmented routing materially affects inclusion time. We further show that execution quality varies substantially across candidate blocks: for the same swap, both execution probability and execution price differ across proposed blocks. To study these differences, we examine competition between two arbitrage bots trading between decentralized and centralized exchanges. We find that, conditional on inclusion in a block that also contains transactions from these bots, user swaps in the same (opposite) direction are less likely (more likely) to execute and receive worse (better) prices. These results show that routing and block composition are central determinants of execution quality and market quality in on-chain markets.
Tech M&A Deals: Palo Alto Networks, Carasent, Euclid - InfotechLead
Technology companies such as Palo Alto Networks, Carasent, Euclid, among others, announced tech M&A deals across AI security platforms
AI Startup Faces Legal Battle Over Unauthorized 'This Is Fine' Meme Use in Ad Campaign
Artisan's use of the 'This Is Fine' meme in an AI sales tool ad has prompted creator Green to seek legal action, claiming unauthorized commercial use.
AI Dispatch: Daily Trends and Innovations – May 4, 2026 | Mark Cuban, Nvidia, ChatGPT Images, Mayo Clinic, and HUMAIN ONE
Artificial intelligence is no longer moving through the market as a single story. Contents Mark Cuban’s warning: AI is coming for tasks first, then roles Nvidia and China: export policy, market share, and the cost of strategic decoupling Deepfakes and fraud: the synthetic media problem is ...
Opinion | Competition for Pentagon AI contracts ensures troops get best tools - The Washington Post
Silicon Valley has been too ambivalent about American power for the first quarter of this century. But deals announced Friday by the Pentagon show how that’s changing for the better, aided by intense competition within the software industry.
Token Arena: A Continuous Benchmark Unifying Energy and Cognition in AI Inference
arXiv:2605.00300v1 Announce Type: new Abstract: Public inference benchmarks compare AI systems at the model and provider level, but the unit at which deployment decisions are actually made is the endpoint: the (provider, model, stock-keeping-unit) tuple at which a specific quantization, decoding strategy, region, and serving stack is exposed. We introduce TokenArena, a continuous benchmark that measures inference at endpoint granularity along five core axes (output speed, time to first token, workload-blended price, effective context, and quality on the live endpoint) and synthesizes them, together with a modeled energy estimate, into three headline composites: joules per correct answer, dollars per correct answer, and endpoint fidelity (output-distribution similarity to a first-party reference). The framework's novelty is empirical and methodological. Across 78 endpoints serving 12 model families, the same model on different endpoints differs in mean accuracy by up to 12.5 points on math and code, in fingerprint similarity to first party by up to 12 points, in tail latency by an order of magnitude, and in modeled joules per correct answer by a factor of 6.2. We further show that workload-aware blended pricing reorders the leaderboard substantially: 7 of 10 top-ranked endpoints under the chat preset (3:1 input:output) fall out of the top 10 under the retrieval-augmented preset (20:1), and the reasoning preset (1:5) elevates frontier closed models that the chat preset penalizes on price. We release the framework, schema, probe and eval harness, and a v1.0 leaderboard snapshot under CC BY 4.0. TokenArena is a methodology, not a single ranking; we publish full provenance and limitations and welcome external replication.
💸 AI Coding Bills Hit $900/Month; Suddenly Hiring Humans Looks Cheap | by Tom Smykowski | May, 2026 | Medium
The days when $20 could power serious development work are basically over. According to Business Insider, Anthropic estimates that an enterprise developer spends around $900 per month on their AI tools.
Tailoring AI solutions for health care needs | MIT Technology Review
The AI market is full of big promises of grand transformation. Health care is a prime target for those promises, beset as it is by financial pressures, labor shortages, and the growing burden of caring for an aging population. AI developers are targeting functions that vary widely, from curing ...
As Formula One evolves, AI becomes part of the race | Reuters
AI has been innovative in sifting through administrative tasks and interpreting key rules within sporting and technical regulations, helping engineers take swifter decisions during on-track situations which were impossible decades ago.
Healthcare’s AI Agents Aim to Give Doctors Time Back | PYMNTS.com
Healthcare’s next AI test will be whether agents can give doctors and nurses back something far more valuable: time. Across new reports and commentary,
From Burnt-Out Freelancer to Productized Operator (With AI Doing the Boring Bits)
AI works best when your service is already productized.
DOT’s Neil Chaudhry on AI-powered productivity in government | FedScoop
The votes are in for the 2025 FedScoop · Neil Chaudhry, Senior Advisor of AI for the U.S. Department of Transportation, discussed the practical applications of AI in federal workforce reimagining. He emphasized leveraging AI for tasks that improve job satisfaction, such as creating personalized ...
Instance-Aware Parameter Configuration in Bilevel Late Acceptance Hill Climbing for the Electric Capacitated Vehicle Routing Problem
arXiv:2605.00572v1 Announce Type: new Abstract: Algorithm performance in combinatorial optimization is highly sensitive to parameter settings, while a single globally tuned configuration often fails to exploit the heterogeneity of instances. This limitation is particularly evident in the Electric Capacitated Vehicle Routing Problem, where instances differ in structure, demand patterns, and energy constraints. This paper investigates instance-aware parameter configuration for Bilevel Late Acceptance Hill Climbing, a state-of-the-art metaheuristic for the Electric Capacitated Vehicle Routing Problem. An offline tuning procedure is used to obtain instance-specific parameter labels, which are then mapped from instance features via a regression model to enable parameter prediction for unseen instances prior to execution. Experimental results on the IEEE WCCI 2020 benchmark and its extensions show that the proposed approach achieves an average objective value reduction of $0.28\%$ across eight held-out test instances relative to a globally tuned configuration. This corresponds to a significant cost reduction in multimillion-dollar transportation operations.
AIwire - Covering Scientific & Technical AI
SUNNYVALE, Calif., May 4, 2026 — Cerebras Systems, Inc. today announced that it plans to commence the roadshow for its proposed initial public offering of its Class A common stock. Cerebras […]
Physical AI Startup Statistics
Physical AI startup statistics for 2026: robotics funding, robot foundation models, industrial robots, warehouses, drones, autonomous systems, and founder opportunity.
Labor, Society & Culture
On the Role of Artificial Intelligence in Human-Machine Symbiosis
arXiv:2605.00440v1 Announce Type: new Abstract: The evolution of artificial intelligence (AI) has rendered the boundary between humanity and computational machinery increasingly ambiguous. In the presence of more interwoven relationships within human-machine symbiosis, the very notion of AI-generated information becomes difficult to define, as such information arises not from either humans or machines in isolation, but from their mutual shaping. Therefore, a more pertinent question lies not merely in whether AI has participated, but in how it has participated. In general, the role assumed by AI is often specified, either implicitly or explicitly, in the input prompt, yet becomes less apparent or altogether unobservable when the generated content alone is available. Once detached from the dialogue context, the functional role may no longer be traceable. This study considers the problem of tracing the functional role played by AI in natural language generation. A methodology is proposed to infer the latent role specified by the prompt, embed this role into the content during the probabilistic generation process and subsequently recover the nature of AI participation from the resulting text. Experimentation is conducted under a representative scenario in which AI acts either as an assistive agent that edits human-written content or as a creative agent that generates new content from a brief concept. The experimental results support the validity of the proposed methodology in terms of discrimination between roles, robustness against perturbations and preservation of linguistic quality. We envision that this study may contribute to future research on the ethics of AI with regard to whether AI has been used fairly, transparently and appropriately.
Computational Challenges in Scaling Democratic Deliberation
arXiv:2605.01525v1 Announce Type: new Abstract: The paper provides an overview of core functionalities that digital democracy software needs to provide in order to support democratic deliberative processes at scale. Developing these functionalities poses novel computational challenges and requires algorithmic solutions to interesting mathematical problems. The aim of the paper is to break the first ground towards a structured inventory of such problems, and to position possible approaches to them within current academic research in computer science and artificial intelligence.
Remote work expands pathways to upward career mobility
arXiv:2605.01268v1 Announce Type: new Abstract: Geographic constraints have long structured access to high-growth career opportunities, concentrating upward mobility within a limited set of cities and organizations. The expansion of remote work potentially alters this opportunity structure by decoupling job matching from physical proximity, yet its implications for career mobility remain unclear. Using 48 million U.S. job transitions between 2020 and 2024 linked to employer-level measures of remote eligibility, we estimate how entering remote-eligible jobs shapes career outcomes at job transitions. Workers entering remote-eligible jobs experience significantly higher wage growth and higher rates of upward seniority mobility than comparable workers entering fully on-site roles. These transitions are also associated with greater cross-metropolitan job mobility and moves toward smaller, less prestigious employers. Importantly, effects are largest among lower-income workers and those originating from regions with limited high-skill opportunity density. Together, the findings indicate that remote work relaxes geographic constraints in job matching, reshaping the distribution of upward mobility across places and workers.
What Jobs Can AI Learn? Measuring Exposure by Reinforcement Learning
arXiv:2605.02598v1 Announce Type: new Abstract: Which jobs can AI learn to do? We examine this for every occupation in the US economy. Existing indices measure the overlap between AI capabilities and occupational tasks rather than which tasks AI systems can learn to perform, and as a result misclassify occupations where the gap between present capability and learnability is large. Reinforcement learning in post-training, now the dominant paradigm at the frontier, is structured around task completion and maps more directly onto the task-based architecture of occupational classifications than prior approaches. Using LLM annotators guided by a rubric developed with RL experts and validated against confirmed deployment cases, we score all 17,951 ONET tasks for training feasibility and aggregate to the occupation level, producing an RL Feasibility Index. The index diverges sharply from existing AI exposure measures for specific occupation groups: power plant operators, railroad conductors, and aircraft cargo handling supervisors score high on RL feasibility but low on general AI exposure, while creative and interpersonal roles (musicians, physicians, natural sciences managers) show the reverse. These divergences carry direct implications for policy interventions.
Generative-AI and the transformation of workforce. A job postings-driven analysis
arXiv:2605.00843v1 Announce Type: new Abstract: This paper investigates how generative-artificial intelligence AI is reshaping job requirements, skill compositions and sectoral dynamics across global labor markets. It examines the evolving frequency and framing of AI-related competencies in job postings, exploring whether generative-AI functions primarily as an augmentative or substitutive force in the workplace. A large-scale, multi-source corpus of over 150,000 English-language job postings 2018-2025 is compiled from twelve open-access datasets and one public API. The analytical framework integrates lexical skill extraction, semantic framing, topic modeling, BERTopic, LDA, KMeans, and time-series forecasting ARIMA. Skill mentions are categorized into five dimensions: AI_Data, Routine, Soft_Meta, Domain_Specific and Leadership, while cross sectoral analyses and correlation matrices quantify interdependencies between competencies. Sentence-transformer embeddings and cosine similarity are used to compute a Framing Index, distinguishing augmentation- versus automation-oriented discourse. Investigating job postings, our research contributes a replicable, data driven methodology for mapping the diffusion of AI related skills across industries and time. Results reveal a sharp post-2021 increase in AI-related skill mentions: prompt engineering, fine-tuning and model validation, accompanied by a decline in routine tasks: data entry and manual coding. Forecasts suggest sustained growth in AI_Data and Soft_Meta skills through 2025, signaling a structural convergence toward hybrid human-AI expertise as a new foundation of employability.
Recruiters turn to AI in quest to find the perfect connection
Technology is being used to ‘clear the decks for human moments’
BofA throws cold water on AI apocalypse panic: 60% of today’s jobs didn’t exist in 1940
AI will reshape 840 million jobs, BofA says. That’s not the same as destroying them.
The Real Job Destruction from AI Is Hitting Before Careers Can Start | Yale Insights
Yale’s Jeffrey Sonnenfeld and his co-authors say that the impact of AI can be seen among recent college graduates, who are finding it harder and harder to get that first job. With no entry to the workforce, how will younger people develop the skills and wisdom to lead in the future?
Gallup finds AI not eliminating creative jobs despite exposure fears | Fox Business
AI and creative jobs may coexist better than feared. A Gallup analysis finds little evidence that generative AI has broadly reduced earnings for artists or caused widespread job losses.
Technology & Infrastructure
Organizational Theory Is Essential for Managing the Complexity of Multi-Agent AI Systems
Current approaches to agentic systems rely too heavily on technical control planes, ignoring the necessary organizational frameworks. Effective deployment requires integrating management principles like decision rights and spans of control.
TADI: Tool-Augmented Drilling Intelligence via Agentic LLM Orchestration over Heterogeneous Wellsite Data
arXiv:2605.00060v1 Announce Type: new Abstract: We present TADI (Tool-Augmented Drilling Intelligence), an agentic AI system that transforms drilling operational data into evidence-based analytical intelligence. Applied to the Equinor Volve Field dataset, TADI integrates 1,759 daily drilling reports, selected WITSML real-time objects, 15,634 production records, formation tops, and perforations into a dual-store architecture: DuckDB for structured queries over 12 tables with 65,447 rows, and ChromaDB for semantic search over 36,709 embedded documents. Twelve domain-specialized tools, orchestrated by a large language model via iterative function calling, support multi-step evidence gathering that cross-references structured drilling measurements with daily report narratives. The system parses all 1,759 DDR XML files with zero errors, handles three incompatible well naming conventions, and is backed by 95 automated tests plus a 130-question stress-question taxonomy spanning six operational categories. We formalize the agent's behavior as a sequential tool-selection problem and propose the Evidence Grounding Score (EGS) as a simple grounding-compliance proxy based on measurements, attributed DDR quotations, and required answer sections. The complete 6,084-line, framework-free implementation is reproducible given the public Volve download and an API key, and the case studies and qualitative ablation analysis suggest that domain-specialized tool design, rather than model scale alone, is the primary driver of analytical quality in technical operations.
AI Agents for Sustainable SMEs: A Green ESG Assessment Framework
arXiv:2605.00841v1 Announce Type: cross Abstract: This study presents a novel, AI-driven framework for assessing Environmental, Social, and Governance (ESG) performance in European small and medium-sized enterprises (SMEs). An initial phase established expert-validated ESG baseline scores from a subset of the Flash Eurobarometer FL549 survey data. In the second phase, a scalable AI agent system, built on the n8n automation platform, applied these baselines to perform automated ESG classification and generate contextual recommendations using large language models (LLMs). The results demonstrate the AI system's high consistency with human-derived outputs, thereby supporting more effective monitoring and intervention strategies aligned with the European Green Deal.
Building an agentic AI strategy that pays off - without risking business failure | ZDNET
Companies are chasing tenfold AI gains, but many projects are failing fast. We break down the real risks and show you how to turn agentic AI into reliable, profitable outcomes.
Payment Networks Ready Infrastructure for Agentic Commerce at Scale | PYMNTS.com
Payment networks are moving agentic commerce from pilots into payment environments, using existing credentials and acceptance infrastructure.
Agentic AI’s Impact on Commercial Real Estate Goes Beyond Time Saved
The potential impact extends well beyond individual workflows, our guest columnist writes.
You built it with AI. Now run it with AI.
The next generation of companies will be designed before they're staffed.
Agentic Coding Is a Trap
A critical look at the current trend of agentic coding and why it may be problematic for developers.
AI agents for SMBs - by Robin Lowe - The Lowe-Down
At its core, an AI agent is software that can plan, make decisions, and take action toward a goal with limited supervision.
Lessons from the agentic AI trailblazers
Many businesses have yet to deploy agents, so what tips can early adopters offer?
AgentReputation: A Decentralized Agentic AI Reputation Framework
arXiv:2605.00073v1 Announce Type: new Abstract: Decentralized, agentic AI marketplaces are rapidly emerging to support software engineering tasks such as debugging, patch generation, and security auditing, often operating without centralized oversight. However, existing reputation mechanisms fail in this setting for three fundamental reasons: agents can strategically optimize against evaluation procedures; demonstrated competence does not reliably transfer across heterogeneous task contexts; and verification rigor varies widely, from lightweight automated checks to costly expert review. Current approaches to reputation drawing on federated learning, blockchain-based AI platforms, and large language model safety research are unable to address these challenges in combination. We therefore propose \textbf{AgentReputation}, a decentralized, three-layer reputation framework for agentic AI systems. The framework separates task execution, reputation services, and tamper-proof persistence to both leverage their respective strengths and enable independent evolution. The framework introduces explicit verification regimes linked to agent reputation metadata, as well as context-conditioned reputation cards that prevent reputation conflation across domains and task types. In addition, AgentReputation provides a decision-facing policy engine that supports resource allocation, access control, and adaptive verification escalation based on risk and uncertainty. Building on this framework, we outline several future research directions, including the development of verification ontologies, methods for quantifying verification strength, privacy-preserving evidence mechanisms, cold-start reputation bootstrapping, and defenses against adversarial manipulation.
AEM: Adaptive Entropy Modulation for Multi-Turn Agentic Reinforcement Learning
arXiv:2605.00425v1 Announce Type: new Abstract: Reinforcement learning (RL) has significantly advanced the ability of large language model (LLM) agents to interact with environments and solve multi-turn tasks. Yet effective training remains challenging, as sparse, outcome-only rewards make it difficult to assign credit to individual steps in an agent's action trajectory. A common remedy is to introduce dense intermediate supervision, such as process reward models or auxiliary self-supervised signals, but this increases supervision and tuning complexity and often generalizes poorly across tasks and domains. This paper presents AEM, a supervision-free credit assignment method that adaptively modulates entropy dynamics during RL training to achieve a more effective exploration-exploitation trade-off. Theoretically, we elevate entropy analysis from the token level to the response level to reduce token sampling variance and show that entropy drift under natural gradients is intrinsically governed by the product of the advantage and the relative response surprisal. Specifically, we derive a practical proxy to reshape training dynamics, enabling a natural transition from exploration to exploitation. Extensive experiments across various benchmarks and models ranging from 1.5B to 32B parameters demonstrate the effectiveness of AEM, including a notable 1.4 percent gain when integrated into a state-of-the-art baseline on the highly challenging SWE-bench-Verified benchmark.
Agentic AI, fraud and the fight for customer loyalty | Retail Dive
As AI-powered commerce grows into a multi-trillion-dollar space, speed, personalization, and automation are becoming standard expectations. However, this shift fundamentally changes how security must be approached. When bots act on behalf of users, businesses are no longer verifying just a person. They must also verify the legitimacy and authority of the agent acting for them. This introduces a new layer of complexity that many current systems ...
Single Agent vs Multi-Agent: When to Build a Multi-Agent System
A practical guide to understanding AI agent design, ReAct workflows, and when to scale from a single agent to a multi-agent system.
Market Power and Distributed Solar Integration in Microgrids under Limited Regulation
arXiv:2603.16893v2 Announce Type: replace-cross Abstract: Decentralized electricity systems increasingly emerge where centralized grids fail to provide reliable supply. In such settings, privately operated neighborhood microgrids, often based on diesel generators, exhibit significant market power, limited regulatory oversight, and high environmental externalities. In parallel, households increasingly deploy off-grid solar photovoltaic (PV) systems to gain control over electricity supply. However, these systems suffer from curtailed excess generation during peak solar hours and unreliable access at other times. While prior studies have optimized microgrids in low-reliability grid contexts from a techno-economic perspective, they largely neglect the market power exerted by monopolistic private generators. This paper addresses this gap by developing a bi-level game-theoretic model that enables household-generated electricity to be fed into the microgrid while explicitly accounting for the market power of a neighborhood diesel generator company (DGC). The regulator sets price and feed-in-tariff caps to maximize household economic surplus (HES), while the DGC acts as a profit-maximizing agent controlling access and supply. The model is illustrated using high-resolution empirical data from Lebanon. Results show that: (i) price and feed-in-tariff caps substantially increase HES and consistently induce significant household PV feed-in to the microgrid; (ii) higher DGC budgets or greater PV-owner penetration lead to pronounced gains in HES; and (iii) the renewable energy share reaches 60% under base conditions and approaches 100% at sufficiently high budgets or PV-owner penetration levels, compared to 0% under the status quo.
The Hidden Cost of Thinking: Energy Use and Environmental Impact of LMs Beyond Pretraining
arXiv:2605.01158v1 Announce Type: new Abstract: Modern language model development extends far beyond pretraining, yet environmental reporting remains narrowly focused on the cost of training a single final model. In this work, we provide the first detailed breakdown of the environmental impact of a full model development pipeline, from pretraining through supervised fine-tuning, preference optimization, and reinforcement learning, for Olmo 3, a family of 7 billion and 32 billion parameter models in both instruction-following and reasoning variants. We find that reasoning models are 17x more expensive to post-train than their instruction-tuned counterparts in terms of datacenter energy, driven by reinforcement learning rollout generation. Development costs (including experimentation, failed runs, and ablations) account for 82.2% of total compute, a roughly 65% increase over the ~50% reported for pretraining-focused pipelines in prior work. In total, we estimate our model development process consumed ~12.3 GWh of datacenter energy, emitted 4,251 tCO2eq, and consumed 15,887 kL of water, with water consumption driven entirely by power generation infrastructure rather than data center cooling. These costs, which are almost entirely unreported by model developers, are growing rapidly as post-training pipelines become more complex, and must be accounted for in environmental reporting standards and by the research community working to reduce AI's environmental impact.
Hugging Carbon: Quantifying the Training Carbon Emissions of AI Models at Scale
arXiv:2605.01549v1 Announce Type: new Abstract: The scaling-law era has transformed artificial intelligence from research into a global industry, but its rapid growth raises concerns over energy usage, carbon emissions, and environmental sustainability. Unlike traditional sectors, the AI industry still lacks systematic carbon accounting methods that support large-scale estimates without reproducing the original model. This leaves open questions about how large the problem is today and how large it might be in the near future. Given that the Hugging Face (HF) platform well represents the broader open-source community, we treat it as a large-scale, publicly accessible, and audit-ready corpus for carbon accounting. We propose a FLOPs-based framework to estimate aggregate training emissions of HF open-source models. Considering their uneven disclosure quality, we introduce a tiered approach to handle incomplete metadata, supported by empirical regressions that verify the statistical significance. Compute is also converted to AI training carbon intensity (ATCI, emissions per compute), a metric to assess the sustainability efficiency of model training. Our results show that training the most popular open-source models (with over 5,000 downloads) has resulted in approximately $5.8\times10^4$ metric tons of carbon emissions. This paper provides a scalable framework for emission estimations and a practical methodology to guide future standards and sustainability strategies in the AI industry.
How a Single Piece of Heavy Machinery Is Reshaping the AI Race - The National Interest
Tags: China, Data Centers, European ... Mitsubishi Heavy Industries, Siemens Energy, and United States ... AI’s next bottleneck is power, with turbine shortages and grid constraints reshaping where and how fast data center infrastructure can scale globally....
The AI electric grid calls for a mixed‑fleet transmission strategy | Utility Dive
In the AI-era planning environment, hybridization is not a compromise — it is the engineering approach best aligned with real world system needs, timelines and locational realities, writes NYISO planning engineer Anees Jeddy.
AI demand lifts silicon wafer shipments 13.1% in Q1
SEMI says global silicon wafer shipments rose 13.1% in Q1, as AI data-centre demand for advanced logic, memory and power devices stayed strong.
Brookfield Infrastructure Q1 2026 Earnings: FFO Hits $709M, AI Data Center Investments Drive 10% Growth - News and Statistics - IndexBox
Brookfield Infrastructure’s Q1 2026 FFO rose 10% to $709M, accelerating from 6% a year ago, fueled by AI data center build-out, new fiber and pipeline acquisitions, and high-end organic growth. The 4.9% dividend remains supported by a strong growth pipeline.
What Was Discussed at Google’s White House Meeting About A.I.
The technology giant met with administration officials last week to address a growing concern in Washington: insufficient computing power for artificial intelligence.
Peter Thiel backs $1bn ocean data centre start-up powered by waves
US tech billionaire leads $140mn investment into Panthalassa as search for AI power pushes into exotic new frontiers
Nvidia's exposure to Asian supply chains for components hits 90% of its production costs — marked increase from 65% could intensify as physical AI adds even more exposure | Tom's Hardware
Memory shortages are killing off older Jetson modules while newer ones compete for the same constrained supply.
Analyst Warns Google's Intel TPU Bet Hinges on a Brutal Yield Jump From 90% to 98%
Analyst says Intel's EMIB-T packaging looks promising, but the final 8% climb is harder than going from zero to 90.
The data advantage: How data value is reshaping Southeast Asia's next leap in AI infrastructure - TNGlobal
The next phase of AI growth in Southeast Asia will not be defined by faster algorithms alone. It will depend on whether organizations treat infrastructure as a strategic foundation rather than a backend technical consideration.
Is there "Secret Sauce'' in Large Language Model Development?
arXiv:2602.07238v2 Announce Type: replace-cross Abstract: Do leading LLM developers possess a proprietary ``secret sauce'', or is LLM performance driven by scaling up compute? Using training and benchmark data for 809 models released between 2022 and 2025, we estimate scaling-law regressions with release-date and developer fixed effects. We find clear evidence of developer-specific efficiency advantages, but their importance depends on where models lie in the performance distribution. At the frontier, 80-90% of performance differences are explained by higher training compute, implying that scale--not proprietary technology--drives frontier advances. Away from the frontier, however, proprietary techniques and shared algorithmic progress substantially reduce the compute required to reach fixed capability thresholds. Some companies can systematically produce smaller models more efficiently. Strikingly, we also find substantial variation of model efficiency within companies; a firm can train two models with more than 40x compute efficiency difference. We also discuss the implications for AI leadership and capability diffusion.
TUR-DPO: Topology- and Uncertainty-Aware Direct Preference Optimization
arXiv:2605.00224v1 Announce Type: new Abstract: Aligning large language models (LLMs) with human preferences is commonly done via reinforcement learning from human feedback (RLHF) with Proximal Policy Optimization (PPO) or, more simply, via Direct Preference Optimization (DPO). While DPO is stable and RL-free, it treats preferences as flat winner vs. loser signals and is sensitive to noisy or brittle preferences arising from fragile chains of thought. We propose TUR-DPO, a topology- and uncertainty-aware variant of DPO that rewards how answers are derived, not only what they say, by eliciting lightweight reasoning topologies and combining semantic faithfulness, utility, and topology quality into a calibrated uncertainty signal. A small learnable reward is factorized over these signals and incorporated into an uncertainty-weighted DPO objective that remains RL-free and relies only on a fixed or moving reference policy. Empirically, across open 7-8B models and benchmarks spanning mathematical reasoning, factual question answering, summarization, and helpful/harmless dialogue, TUR-DPO improves judge win-rates, faithfulness, and calibration relative to DPO while preserving training simplicity and avoiding online rollouts. We further observe consistent gains in multimodal and long-context settings, and show that TUR-DPO matches or exceeds PPO on reasoning-centric tasks while maintaining operational simplicity.
Thinking in Text and Images: Interleaved Vision--Language Reasoning Traces for Long-Horizon Robot Manipulation
arXiv:2605.00438v1 Announce Type: new Abstract: Long-horizon robotic manipulation requires plans that are both logically coherent and geometrically grounded. Existing Vision-Language-Action policies usually hide planning in latent states or expose only one modality: text-only chain-of-thought encodes causal order but misses spatial constraints, while visual prediction provides geometric cues but often remains local and semantically underconstrained. We introduce Interleaved Vision--Language Reasoning (IVLR), a policy framework built around \trace{}, an explicit intermediate representation that alternates textual subgoals with visual keyframes over the full task horizon. At test time, a single native multimodal transformer self-generates this global semantic-geometric trace from the initial observation and instruction, caches it, and conditions a closed-loop action decoder on the trace, original instruction, and current observation. Because standard robot datasets lack such traces, we construct pseudo-supervision by temporally segmenting demonstrations and captioning each stage with a vision-language model. Across simulated benchmarks for long-horizon manipulation and visual distribution shift, \method{} reaches 95.5\% average success on LIBERO, including 92.4\% on LIBERO-Long, and 59.4\% overall success on SimplerEnv-WidowX. Ablations show that both modalities are necessary: without traces, LIBERO-Long success drops to 37.7\%; text-only and vision-only traces reach 62.0\% and 68.4\%, while the full interleaved trace reaches 92.4\%. Stress tests with execution perturbations and masked trace content show moderate degradation, suggesting that the trace can tolerate local corruption and moderate execution drift, but remains limited under stale or incorrect global plans.
LLM-based uncertainty assessment of social media situational signals for crisis reporting
arXiv:2605.00829v1 Announce Type: new Abstract: Social media has become a critical source of situational awareness during disasters, providing real-time insights into evolving impacts and emerging needs. To support crisis response at scale, recent work has increasingly leveraged large language models (LLMs) to automatically classify and summarize situational information from social media streams. However, existing approaches implicitly assume that extracted situational claims are equally plausible, despite information quality varying substantially as a crisis unfolds. In this work, we propose an uncertainty-aware framework for automated situational awareness reporting that explicitly accounts for the plausibility of social media claims. First, we classify social media posts according to an established situational awareness schema. Second, we introduce an uncertainty assessment layer that evaluates whether individual situational claims plausibly reflect real-world conditions when conditioned on external proxy data, while explicitly eliciting the model's confidence in this judgment. Third, we use these uncertainty assessments to generate crisis reports that communicate not only what is being reported, but how certain those reports are. We apply this framework to over 200,000 earthquake-related Twitter/X posts, using impact summaries from the USGS PAGER as a representative external proxy. We argue that explicitly representing uncertainty supports human crisis communicators in prioritizing information under time pressure, and provides a framework for integrating external proxy data into LLM-based situational awareness pipelines.
Anthropic Leadership and the Transparency of Public AI Research and Disclosure
The reliance of AI leadership on public sources for commentary highlights a gap between internal proprietary knowledge and public disclosure. This dynamic influences how industry experts shape market expectations.
Why data centers now belong on the critical infrastructure list | CyberScoop
As AI shifts data centers from digital backbones to strategic targets, industry experts argue they must officially be classified as critical infrastructure. Learn why rising physical and cyber threats are making operational resilience a board-level national security priority.
US DoD announces agreements with frontier AI companies for classified work
The US Department of Defense announced deals with seven AI companies to deploy advanced capabilities on classified networks, even as Anthropic challenges its supply chain risk designation.
Army convenes industry leaders for AI tabletop exercise focused on cyber defense | Article | The United States Army
WASHINGTON — On April 27, the Army convened 14 senior cybersecurity executives from leading technology companies at the Pentagon for the second iteratio...
Shadow IT has given way to shadow AI. Enter AI-BOMs
'If you don't have visibility, you can't understand what to protect' When it comes to securing enterprise supply chains, now heavily infused with AI applications and agents, a software bill of materials (SBOM) no longer provides a complete inventory of all the components in the environment. Enter AI-BOMs.…
Minimal, Local, Causal Explanations for Jailbreak Success in Large Language Models
arXiv:2605.00123v1 Announce Type: new Abstract: Safety trained large language models (LLMs) can often be induced to answer harmful requests through jailbreak prompts. Because we lack a robust understanding of why LLMs are susceptible to jailbreaks, future frontier models operating more autonomously in higher-stakes settings may similarly be vulnerable to such attacks. Prior work has studied jailbreak success by examining the model's intermediate representations, identifying directions in this space that causally encode concepts like harmfulness and refusal. Then, they globally explain all jailbreak attacks as attempting to reduce or strengthen these concepts (e.g., reduce harmfulness). However, different jailbreak strategies may succeed by strengthening or suppressing different intermediate concepts, and the same jailbreak strategy may not work for different harmful request categories (e.g., violence vs. cyberattack); thus, we seek to give a local explanation -- i.e., why did this specific jailbreak succeed? To address this gap, we introduce LOCA, a method that gives Local, CAusal explanations of jailbreak success by identifying a minimal set of interpretable, intermediate representation changes that causally induce model refusal on an otherwise successful jailbreak request. We evaluate LOCA on harmful original-jailbreak pairs from a large jailbreak benchmark across Gemma and Llama chat models, comparing against prior methods adapted to this setting. LOCA can successfully induce refusal by making, on average, six interpretable changes; prior work routinely fails to achieve refusal even after 20 changes. LOCA is a step toward mechanistic, local explanations of jailbreak success in LLMs. Code to be released.
Google made agentic AI governance a product. Enterprises still have to catch up.
Agentic systems multiply identities and permissions at a pace that traditional human-centric identity and access management models were never built to handle. Once agents start acting across systems, the governance question shifts from which model is approved to what actions a given agent can ...
CISA and partners release agentic AI security guidance to protect critical infrastructure, outline mitigation action - Industrial Cyber
CISA and partners release agentic AI security guidance to protect critical infrastructure, outline mitigation action
Cisco Unveils Open-Source Toolkit to Bolster AI Model Security and Provenance
Cisco has unveiled the open-source Model Provenance Kit to bolster AI model security by tracking their origin and integrity.
AI for Security Infrastructure: Rebalancing Cybersecurity for the Decade Ahead - Security Boulevard
An exploration of the shift from reactive "assume breach" mentalities to AI-driven prevention, highlighting how Domain-Specific Language Models (DSLMs) empower security architects to eliminate configuration drift and tool sprawl.
US government warns of severe CopyFail bug affecting major versions of Linux
Federal authorities have issued a warning regarding a critical vulnerability known as 'CopyFail' that impacts several versions of the Linux operating system.
Dailypioneer
In fact, in a recent report, Gartner® ... and preemptive cybersecurity capabilities such as predictive threat intelligence, automated moving target defense (AMTD), and advanced cyber deception are needed to defend against AI-driven threats.” ¹ Preemptive risk assessment and system ...
Adoption, Deployment & Impact
Restaurants lean on AI to cut waste and reduce costs
Hospitality sector has been slower than others at adopting the technology
Permission-Aware AI Repository Gateway Market to Reach USD 6.68 Billion by 2036 as Secure Enterprise AI Retrieval and Governance Drive Adoption | Morningstar
Market Value Analysis: Secure AI Retrieval Becomes Core Enterprise Infrastructure · Between 2026 and 2030, enterprise adoption of permission-aware AI repository gateways is expected to accelerate as organizations scale internal copilots, AI assistants, and knowledge agents connected to live ...
Integrated Digital Management System for Railway Workshops: A Modular Multi-Workflow Architecture for Machine, Permit, Contract, and Incident Management
arXiv:2605.00840v1 Announce Type: new Abstract: Indian Railway workshops form a critical component of rolling stock maintenance infrastructure, employing more than 2.5 lakh personnel across 44 major workshops nationwide. However, safety management in many workshops still relies on fragmented manual processes, resulting in delayed approvals, incomplete documentation, and increased exposure to operational hazards. Field safety observations indicate that lacerations (28.7%) and abrasions (21%) remain among the most frequent workplace injuries, highlighting the need for structured digital safety workflows. This paper presents the Integrated Digital Management System for Railway Workshops, a modular multi-workflow digital platform developed to improve safety governance and workflow transparency. The proposed system integrates four primary modules: Machine and Plant Management, Permit-to-Work (PTW) Management, Contract Management, and Incident Management. The Permit-to-Work module digitizes hazardous work authorization in accordance with IS 17893:2022, while the Contract Management module supports workforce validation and regulatory oversight. The Incident Management module enables rapid reporting, investigation tracking, and corrective action workflows. Functional evaluation in a railway workshop-oriented deployment scenario demonstrated measurable operational improvements, including a reduction in permit processing time by approximately 35%, improved incident reporting response time by nearly 40%, and enhanced workflow traceability across multiple safety modules. The proposed system establishes a scalable foundation for digital safety governance in large-scale railway workshop environments.
National Safety Council and Wolters Kluwer Enablon Report Finds Adoption of AI Growing Rapidly Among EHS Pros | ISHN
Enthusiasm for AI accompanied by caution, as 90% of safety professionals agree guardrails are needed.
Enterprises Are Deploying AI Faster Than They Can Govern: Why Identity Is Becoming Critical
On April 30, 2026, Okta made its Okta for AI Agents platform generally available, bringing to market a new framework designed to help enterprises manage the growing complexity of AI-driven systems. At the core of this launch is Okta’s “secure agentic enterprise” blueprint, which introduces ...
Infor's Technology Tackles Healthcare's AI Execution Gap | Healthcare Digital
Infor's new platform tackles industry-specific AI scaling challenges with robust governance and compliance features for healthcare providers
How AI Tools Generate Technical Debt in IoT Systems — and What to Do About It
AI tools speed up IoT development, but the same code that looks correct can silently break thousands of devices at once.
From ambition to action: Making AI work for SA businesses | ITWeb
Tools like Microsoft 365 Copilot, Azure AI services and Power Platform are not standalone AI experiments; they are designed to integrate with the environments most South African businesses already operate in. That matters because one of the most common barriers to AI deployment is not budget ...
AI in Practice
Water utilities jettison listening sticks; restaurants aim to cut waste and reduce costs; recruiters’ quest to find the perfect connection; start-ups move fast with AI-generated code; hedge funds seek an edge; wealth managers insist AI can work in their favour
Agentic AI for Trip Planning Optimization Application
arXiv:2605.00276v1 Announce Type: new Abstract: Trip planning for intelligent vehicles increasingly requires selecting optimal routes rather than merely producing feasible itineraries, as interacting factors such as travel time, energy consumption, and traffic conditions directly affect plan quality. Yet existing systems are largely designed for feasibility-oriented planning, and current benchmarks provide only reference answers without ground truth, preventing objective evaluation of optimization performance. In our paper, we address these limitations with an agentic AI framework that enables dynamic refinement through an orchestration agent coordinating specialized agents for traffic, charging, and points of interest, and with the Trip-planning Optimization Problems Dataset, which supplies definitive optimal solutions and category-level task structure for fine-grained analysis. Experiments show that our system achieves 77.4\% accuracy on the TOP Benchmark, significantly outperforming single-agent and workflow-based multi-agent baselines, demonstrating the importance of orchestrated agentic reasoning for robust trip planning optimization.
Who Decides What Is Harmful? Content Moderation Policy Through A Multi-Agent Personalised Inference Framework
arXiv:2605.01416v1 Announce Type: new Abstract: The increasing scale and complexity of online platforms raises critical policy questions around harmful content, digital well-being, and user autonomy. Traditional content moderation systems rely on centralised, top-down rules, often failing to accommodate the subjective nature of harm perception. This paper proposes an LLM-based multi-agent personalised inference framework that filters content based on unique sensitivity profiles of individual users. Our architecture combines domain-specific Expert Agents, a Manager Agent for orchestrating content analysis and agent selection, and a Ghost Profile Agent for simulating user perspectives, to inform moderation decisions. Evaluated against a range of non-personalised baselines, the system demonstrates up to a 32% improvement in accuracy, showing increased alignment with individual user sensitivities. Beyond technical performance, our framework provides policy-relevant insights for platform governance, providing a scalable way to reconcile moderation policies with societal and individual digital rights
Singapore central bank pilots cross-bank AI to detect scams earlier
The Monetary Authority of Singapore is collaborating with five banks and government agencies to test AI and machine learning for pre-emptive scam detection using pooled cross-bank transaction data.
Penske Launches AI-Driven Platform for Real-Time Supply Chain Visibility and Efficiency
Penske Logistics has unveiled Supply Chain Insight, a cloud-native platform offering real-time visibility and AI-driven decision-making for supply chain performance.
A-Cube lands €4 million to support e-invoicing and real-time tax reporting expansion from Italy
A-Cube, a Milan startup specialising in automated digital tax compliance, has secured a €4 million investment round to support product expansion, develop solutions adjacent to digital tax reporting, and integrate AI models into its platform. The funding was led by P101 SGR, with participation from Sella Direct Ventures. “Global taxation is undergoing a radical shift, […]
AI finds signs of pancreatic cancer before tumors develop
New AI research shows promise in detecting early signs of pancreatic cancer before physical tumors are even present.
Extracting Contract Insights with PwC’s AI-Driven Annotation on AWS
A use case detailing an AI-driven contract analysis system that extracts structured insights and supports natural language questioning, useful for legal and procurement teams.
Struggling to Keep Up? These 15 Edtech Companies Are Using AI to Give Teachers Their Lives Back
Generative AI is transforming the classroom. From reducing teacher burnout to "AI-proofing" careers, meet the 15 innovative companies reshaping the future of education in 2026.
Principles and Guidelines for Randomized Controlled Trials in AI Evaluation
arXiv:2605.02050v1 Announce Type: new Abstract: This work establishes a foundational framework for standardizing AI evaluation RCTs (sometimes called human uplift studies). Drawing on established experimental practices from disciplines with established RCT traditions, including software engineering, economics, clinical and health sciences, and psychology, we adopt the (Shadish et al., 2002) four-validity framework and extend it with a fifth principle on transparency, repeatability, and verification adapted from the Transparency and Openness Promotion (TOP) Guidelines (Center for Open Science, 2025). We operationalize all five principles into 33 guidelines adapted for AI evaluation RCT contexts, expressed as requirements with rationales, implementation instructions, and evidence bases. We position the principles and guidelines as serving three key roles for AI evaluation RCTs: a design tool for planning studies, an evaluation rubric for assessing existing work, and a blueprint for standard setting as the field converges on norms. Our framework extends prior work by centering evaluation on human performance rather than model output alone, formalizing causal inference through RCT methodology for AI contexts, integrating heterogeneity analysis and practical significance assessment, implementing a graded transparency and repeatability framework, and addressing AI-specific challenges including model versioning, human-AI interaction dynamics, contamination and spillover effects, and equitable impact assessment.
Evaluating AI at Scale: How Thumbtack Approaches Reliability, Safety, and Quality in GenAI
Thumbtack operationalizes GenAI evaluation by treating evals as core infrastructure, incorporating rubrics, AI-as-judge, human review, and production monitoring into their workflow.
A Large-Scale Observational Study on Obtaining Lightweight, Randomized Weekly Student Feedback
arXiv:2605.02281v1 Announce Type: new Abstract: Conventional methods of obtaining student feedback on course experience face a fundamental tradeoff between feedback frequency and quality: as feedback requests become more frequent, participation often declines, and responses become less thoughtful over time. To obtain both timely and thoughtful feedback from students, Kim and Piech (Learning at Scale, 2023) recently proposed a simple, lightweight course feedback mechanism: surveying each student a small number of times per term during randomly selected weeks. Named High-Resolution Course Feedback (HRCF), this method has been shown to elicit feedback that instructors find helpful without imposing excessive burden on students. An important question, however, remains unanswered: is the use of this simple method associated with measurable improvements in students' actual course experiences? We study HRCF use across 103 course offerings, totaling 24,216 student enrollments, over four years from Fall 2021 through Fall 2025, spanning 42 unique computer science courses at an R1 institution. Through a regression analysis of four end-of-term student evaluation items for these courses, we find that first-time use of HRCF is not associated with a measurable change in average student ratings. However, among small- and medium-enrollment (<250 students) course offerings, continued HRCF use is associated with average rating increases of 0.045 to 0.048 points per additional term of use for learning-related items. We observe no statistically significant associations for large-enrollment (250 or more students) course offerings, nor for items measuring instructional quality and course organization. Together, these findings suggest that sustained HRCF use may support improvements in students' learning experiences, but that further design enhancements may be needed to produce measurable improvements in instructional quality and course organization.
A new dashboard tracks how much KPMG workers use AI. They say it's easy to game the system.
KPMG says it hopes the dashboard encourages more "frequent and sophisticated" AI use among its US advisory's 10,000 workers.
The "Astonishing Regularity'' Revisited: Sensitivity of Learning-Rate Estimates to Practice-Sequence Length
arXiv:2605.01690v1 Announce Type: new Abstract: A 2023 \textit{PNAS} study by Koedinger et al. (2023) fit the individual Additive Factors Model (iAFM) to 27 educational datasets and reported an ``astonishing regularity'' in student learning rates: students vary substantially in initial knowledge but learn at remarkably similar rates with practice. We probe a largely unexamined assumption underlying this finding -- that observation length in student log data is ignorable for mixed-effects estimation -- by refitting the iAFM on 26 of the original datasets while systematically truncating practice sequences at various depths, holding the set of students and knowledge components constant. Capping at the first ten opportunities per student-skill pair inflates the median estimated IQR of student learning rates by 75\%; capping at five inflates it by 205\%, with individual datasets ranging from negligible to 17-fold. The magnitude of this sensitivity diverges from what standard estimation theory predicts under ignorable truncation, and the dataset-specific heterogeneity is substantial. Three candidate mechanisms from adjacent literatures could account for the pattern -- informative observation length, functional-form misspecification, and identification weakness from sparse per-pair data -- but observational analysis on these data alone cannot adjudicate among them. We argue that practice sequence length distributions are an unexamined property of mixed-effects estimation on observational learning data, deserving explicit reporting before conclusions about learning-rate heterogeneity are drawn.
How to Move from AI Experimentation to AI Transformation
This article provides a framework for scaling AI beyond fragmented pilots, using a Lowe's case study to illustrate the path toward enterprise-wide AI transformation.
Council Post: It’s Time To Retire The Corporate Training Course. AI Has Already Replaced It
If AI changes how people learn, it also changes how organizations can measure it.
Alvarez & Marsal Wants to Make $3.5 Billion From AI Work by 2028
Alvarez & Marsal wants to generate 50% of its revenue from artificial intelligence work by 2028, which would represent as much as $3.5 billion in earnings.
How process intelligence could decide AI's ROI
This podcast episode explores why process intelligence is the essential foundation for mapping workflows before implementing AI agents to avoid scaling inefficiency.
Geopolitics, Policy & Governance
ARMOR 2025: A Military-Aligned Benchmark for Evaluating Large Language Model Safety Beyond Civilian Contexts
arXiv:2605.00245v1 Announce Type: new Abstract: Large language models (LLMs) are now being explored for defense applications that require reliable and legally compliant decision support. They also hold significant potential to enhance decision making, coordination, and operational efficiency in military contexts. These uses demand evaluation methods that reflect the doctrinal standards that guide real military operations. Existing safety benchmarks focus on general social risks and do not test whether models follow the legal and ethical rules that govern real military operations. To address this gap, we introduce ARMOR 2025, a military aligned safety benchmark grounded in three core military doctrines the Law of War, the Rules of Engagement, and the Joint Ethics Regulation. We extract doctrinal text from these sources and generate multiple choice questions that preserve the intended meaning of each rule. The benchmark is organized through a taxonomy informed by the Observe Orient Decide Act (OODA) decision making framework. This structure enables systematic testing of accuracy and refusal across military relevant decision types. This benchmark features a structured 12-category taxonomy, 519 doctrinally grounded prompts, and rigorous evaluation procedures applied to 21 commercial LLMs. Evaluation results reveal critical gaps in safety alignment for military applications.
Are we Doomed to an AI Race? Why Self-Interest Could Drive Countries Towards a Moratorium on Superintelligence
arXiv:2605.01297v1 Announce Type: new Abstract: This paper uses game theory to argue that, contrary to the prevailing view, a moratorium on Artificial Superintelligence (ASI) can be in a state's self-interest. By formalizing trategic interactions between geopolitical superpowers, we model the trade-off between the benefits of technological supremacy and the catastrophic risks of uncontrolled ASI. The analysis reveals that as the perceived cost of loss of control increases sufficiently relative to other parameters, it becomes in each state's self-interest to impose a moratorium. We further provide empirical evidence suggesting that the global perception of ASI risk is rising, making a stable, rational moratorium increasingly plausible in the current geopolitical landscape.
White House considers government reviews for AI models, NYT reports | Reuters
U.S. President Donald Trump is considering the introduction of government oversight over new models of artificial intelligence, the New York Times reported on Monday, citing officials briefed on the deliberations.
Governing What the EU AI Act Excludes: Accountability for Autonomous AI Agents in Smart City Critical Infrastructure
arXiv:2605.01091v1 Announce Type: new Abstract: When a traffic signal controller adjusts green phases and a grid manager curtails power on the same corridor, each system may comply with its own obligations. The resident who suffers the combined effect has no single authority to hold accountable and, under the EU AI Act, limited means to obtain an explanation. Annex III, point 2 excludes safety-component AI in critical infrastructure from Article 86 explanation rights and Article 27 fundamental-rights impact assessment. Provider and deployer duties under Articles 9-15 still apply, and residual pathways under the GDPR, NIS2, and tortious liability offer partial coverage. The Act's principal resident-facing accountability instruments are nonetheless narrowed for the autonomous infrastructure systems most likely to interact across agencies. The paper traces this accountability deficit through four residual pathways (GDPR Article 22, GDPR transparency obligations, tortious liability, and NIS2) and shows that each is structurally bounded by individual-controller, individual-decision scope. As a governance response, it presents AgentGov-SC, a three-layer architecture (Agent, Orchestration, City) specifying 25 governance measures with bidirectional traceability to the EU AI Act, ISO/IEC 42001, and the NIST AI Risk Management Framework. Five conflict resolution rules and an autonomy-calibrated activation model complete the design. A scenario analysis traces governance activation through a multi-agent corridor cascade involving three documented UAE smart-city systems, with a contrasting single-system scenario confirming proportional activation. The paper contributes a regulatory gap analysis and governance architecture for an increasingly important class of urban AI deployment that existing frameworks treat as bounded and isolated.
The Case for ESM3 as a General-Purpose AI Model with Systemic Risk Under the EU AI Act
arXiv:2605.01611v1 Announce Type: new Abstract: Due to ambiguity in the wording of the EU AI Act, we examine the question of to what extent frontier biological foundation models such as ESM3 are subject to obligations for general-purpose AI models with systemic risk under the EU AI Act. In this paper, we map ESM3 to the biorisk chain, and conclude that it would be desirable if the providers of ESM3 and similar biological models were subject to these obligations, which would require them to assess and mitigate dual-use risks from their models. We then perform an analysis, comparing the attributes of ESM3 to the classification criteria in the AI Act and the supporting material. We conclude that at this time, ESM3 does not appear to be meaningfully regulated by the Act. We then propose remedies to correct the situation.
North Carolina Targets Hyperscale Costs AI Infrastructure Bill
Proposed legislation shifts the burden of power, water, and grid expansion costs onto large data centers.
Connecticut passes social media addiction, chatbot bills
Connecticut lawmakers approved legislation to restrict social media features for minors and establish safety guardrails for AI tools, including chatbots and automated hiring systems.
Executives warn AI risks are outpacing regulation - Business Insurance
Generative artificial intelligence may be transforming business operations at record speed, but regulation and risk management are struggling to keep pace, according to speakers Monday at the Risk &
National security implications in ISDS vis-à-vis AI regulation | White & Case LLP - JDSupra
Technological advances create novel security risks, prompting States to adopt national security measures that restrict foreign investors in this space. As Artificial Intelligence (“AI”) becomes...
Jane Kong - National Security Agency | LinkedIn
The AI Action Plan nods to deepfakes and sets aside money for forensic detection standards, but it falls far short of a full defence. With roughly half of Americans unable to define a deepfake and the nastiest abuses happening well before any judge inspects the evidence, the document amounts ...
Mitigating risk from emerging agentic AI in federal environments | Federal News Network
The challenge is not to halt innovation. It is to ensure that as AI gains agency, agencies retain control and remain protected from fast-evolving threats.
Canada's Privacy Commissioner launches new age assurance guidance to support organizations
Privacy Commissioner Philippe Dufresne launched new guidance on age assurance to help create a safer online experience for children while protecting privacy.
Get the full executive brief
Receive curated insights with practical implications for strategy, operations, and governance.