Sun 19 April 2026
Weekly Brief — Curated and contextualised by Best Practice AI
Stanford Reveals China's AI Parity, TSMC Logs Record Revenues, and IMF Reports Stagnant Productivity
TL;DR Stanford's 2026 AI Index shows China nearly erasing the US lead in model performance and adoption metrics. TSMC reported Q1 2026 revenue of $35.71 billion, up 35% year-over-year, driven by AI chip demand. PwC found 75% of AI's economic value captured by 20% of firms, with leaders prioritizing growth over efficiency. Anthropic's Mythos model sparked warnings from financial officials about vulnerabilities in global banking cyber defenses. Utilities across 51 US investor-owned firms plan $1.4 trillion in spending over five years to support AI data center power needs. Snap laid off 16% of its full-time staff, or about 1,000 employees, as it increases reliance on AI.
Economics & Markets
Why You Should Wait Out AI’s Super-Spending False Start
Merryn Somerset Webb sits down with Janusz Marecki, CEO and founder of Fractal Brain and AI partner at Ahren Innovation Capital, for an insider perspective on artificial intelligence hype versus reality and what may come next. The focus is on large language models (LLMs) and whether they are hitting their fundamental limits. Marecki discusses the data ceiling, diminishing returns from scaling comp
Three-quarters of AI’s economic gains are being captured by just 20% of companies – with the leading companies focused on growth, not just productivity | PwC
A small group of companies is pulling sharply ahead in the race to generate real financial returns from artificial intelligence, according to PwC’s new AI Performance study.
ASML Shares Get Harder to Value as Firm Stops Disclosing Orders
ASML Holding NV will omit the one metric that has driven its share price more than any other when it reports earnings Wednesday, injecting uncertainty into a stock nearing record highs.
Detecting Corporate AI-Washing via Cross-Modal Semantic Inconsistency Learning
arXiv:2604.09644v1 Announce Type: new Abstract: Corporate AI-washing-the strategic misrepresentation of AI capabilities via exaggerated or fabricated cross-channel disclosures-has emerged as a systemic threat to capital market information integrity with the widespread adoption of generative AI. Existing detection methods rely on single-modal text frequency analysis, suffering from vulnerability t
You’re Paying $200/Month for AI. The Same Models Are Now Free.
You’re Paying $200/Month for AI. The Same Models Are Now Free. # Future Digest SubscribeSign in # You’re Paying $200/Month for AI. The Same Models Are Now Free. Apr 14, 2026 ∙ Paid Share Last month, I got my Anthropic invoice. $200. Again. Claude Max 20x. I’ve been on it since January. Before that, I was paying $100 for Max 5x, plus $20 for ChatGPT Plus on the side, plus $20 for Perplexity Pro. Th
Allbirds Soars After Sneaker Firm Rebrands as AI Stock
Allbirds Inc., the once-buzzy maker of wool sneakers valued at more than $4 billion in its heyday, announced a new business plan just days before it planned to close down for good: AI computing infrastructure.
Goldman Sachs' Bordlemay on AI Investment and Inflation Caused by Market Uncertainty
Katherine Bordlemay, Goldman Sachs Asset Management Co-Head of Equity Client Portfolio Management, discussed the current equity market environment at all-time highs. She emphasized that equities are benefiting from upward earnings revisions, supported by significant and durable AI investment spending projected to reach $1 trillion over the next three to four years. (Source: Bloomberg)
a16z’s Martin Casado: It’s not that hard to build AI models
The technologist and investor argues recent progress in AI is an industrial revolution-scale event but warns the ability of the bigger players to raise ‘cheap money’ is time-limited
AI Dominates ETF Flows
Investors are getting picky, and ETFs reveal exactly where the money is going. Seana Smith, Senior Investment Strategist at Global X, joins Bloomberg Open Interest to break down why AI still dominates, how defense stocks are surging on global tensions, and why only a handful of companies will truly win the AI race. (Source: Bloomberg)
Iconiq, Go-To Wealth Adviser for Tech’s Elite, Is Putting Billions Into AI
Last year, Anthropic PBC chief Dario Amodei and a handful of executives traveled 8,000 miles from San Francisco to the Middle East. They were there to meet with some of the world’s most deep-pocketed investors, including Qatar’s sovereign wealth fund and Abu Dhabi-based MGX.
Nvidia rival Cerebras discloses US IPO filing as AI boom drives listings | Reuters
It is focused on inference, the process by which AI systems respond to user queries, and has tied much of its growth to Open AI , including a $20 billion multi-year deal under which the ChatGPT creator will deploy 750 megawatts of Cerebras chips.
Anthropic squeezes enterprises by ejecting bundled tokens from seat deal
Large organizations pushed toward metered pricing.
Meta Expected to Unseat Google as World’s Largest Digital-Ad Player
The owner of Instagram and Facebook has fueled growth with new advertising products and AI.
Is Anthropic 'nerfing' Claude? Users increasingly report performance degradation as leaders push back
A growing number of developers and AI power users are taking to social media to accuse Anthropic of degrading the performance of Claude Opus 4.6 and Claude Code — intentionally or as an outcome of compute limits — arguing that the company’s flagship coding model feels less capable, less reliable and more wasteful with tokens than it did just weeks ago. The complaints have spread quickly on Github
Read OpenAI’s latest internal memo about beating the competition
An internal memo from OpenAI outlines the company's strategy to maintain its competitive edge against rivals like Anthropic.
The 3 forces quietly dismantling the business model that made enterprise software fabulously profitable
Roundtables with senior business leaders in San Francisco and New York revealed a consensus: AI isn't just threatening SaaS, but the entire industry.
Three-quarters of AI’s economic gains are being captured by just 20% of companies – with the leading companies focused on growth, not just productivity | PwC
A small group of companies is pulling sharply ahead in the race to generate real financial returns from artificial intelligence, according to PwC’s new AI Performance study.
Why AI is an operating model shift—Not a technology upgrade - Microsoft in Business Blogs
Why AI is an operating model shift—Not a technology upgrade - Microsoft in Business Blogs April 14, 2026 - Categories - Era of AI more Artificial intelligence isn’t arriving gently in financial services. It’s colliding with an industry defined by scale, trust, and real‑time decision‑making, where speed, security, and reliability are non‑negotiable. For Fiserv, that reality is existential. The comp
Nearly 75pc of AI’s economic value captured by just 20pc of companies
PwC has published a report exploring how only a fraction of organisations are fully capturing the economic value of artificial intelligence.
With labor-force growth slowing to a crawl, what’s the best way to expand the economy? Make the workers you have more productive
An unusual divergence between GDP and new jobs shows worker productivity is making up for a slowdown in immigration and the labor force. Artificial intelligence could help.
AI can power small teams to boost output, write @saraheneedleman in @BusinessInsider.
AI can power small teams to boost output, write @saraheneedleman in @BusinessInsider. https://africa.businessinsider.com/careers/snaps-layoffs-highlight-growing-work-trend-ai-powered-tiny-teams/xw8stkt Here's my quote: "The winners won't simply be the leanest organizations," [Brynjolfsson] said. "They'll be the ones that best redesign work so humans and AI complement each other."
Labor & Society
How will AI change the org chart?
At the micro level, there is evidence that it pays to collapse the layers of command
40% of AI productivity gains lost to rework for errors - CIO
40% of AI productivity gains lost to rework for errors | CIO # 40% of AI productivity gains lost to rework for errors News Analysis Apr 13, 20265 mins ## AI tools are helping organizations get work done faster, but there’s a good chance that your most engaged and talented employees are the ones quietly tasked with the burden of AI-cleanup. Credit: Rob Schultz / Shutterstock While the promise of en
The org chart isn’t ready: How AI exposed the hidden crisis inside the American corporation
A landmark KPMG index finds boards demanding transformation, execs spending billions on tech, and the people caught in the middle burning out.
Snap Is Laying Off 16% of Full-Time Staff as It Embraces A.I.
The company, which owns the social media app Snapchat, said it was laying off about 1,000 employees as it increased its reliance on artificial intelligence.
The $10 Billion Startup Training AI to Replace the White-Collar Workforce
Mercor is promising to replicate most professional work. It was also co-founded by twentysomethings who previously never held a real job.
Alex Imas on Why Economists Might Be Getting AI Wrong
Everyone knows that new technologies can be really disruptive to the labor market, but eventually new jobs emerge and things come back into balance. And there is a sense in which many view AI with the same lens. Yes, there will be pain in some sectors, but then there will be productivity gains and new sources of demand and new opportunities for labor that we can't conceive of yet. But could it be
AI-boosted hacks with Anthropic's Mythos could have dire ... - Reuters
AI-boosted hacks with Anthropic’s Mythos could have dire consequences for banks | Reuters Exclusive news, data and analytics for financial market professionalsLearn more aboutRefinitiv Anthropic logo is seen in this illustration taken March 1, 2026. REUTERS/Dado Ruvic/Illustration/File Photo Purchase Licensing Rights, opens new tab - Summary - Companies - Experts warn Mythos could expose legacy ba
AI agents are scaling faster than cyber defenses
AI agents could soon outnumber humans in the enterprise, necessitating a shift in cybersecurity. Traditional defenses may fail as agent-driven software creation accelerates.
EU Appoints Former Von der Leyen Aide to Top Competition Job
The European Commission has appointed Anthony Whelan, a former aide to President Ursula von der Leyen, to lead its competition policy arm amid growing clashes with the US over regulating Big Tech and a push for merger rules that boost European giants.
Big Tech’s $300mn election war chest rattles Democrats
Pro-industry campaign groups deploy millions amid growing public support for tighter regulation
Why Digital Sovereignty Is About Control, Not Location
Why Digital Sovereignty Is About Control, Not Location Skip to main content We use optional cookies to improve your experience on our websites, such as through social media connections, and to display personalized advertising based on your online activity. If you reject optional cookies, only cookies necessary to provide you the services will be used. You may change your selection by clicking “Man
Anthropic CEO met White House chief of staff as US seeks access to Mythos model
Dario Amodei met with Susie Wiles despite lawsuits over whether AI lab is a national security threat
The U.S. military is missing out because of Hegseth's war on Anthropic
Opinion | Pentagon ban on Anthropic looks shortsighted amid Mythos breakthroughs - The Washington Post Democracy Dies in Darkness Leaving aside the underlying merits of the contractual dispute, it is now abundantly clear that the Trump administration’s aggressive actions against Anthropic have been counterproductive. The cutting-edge artificial intelligence business announced last week that its ne
US 3rd Circuit fair-use opinion could benefit AI companies
A recent fair-use decision in the US Court of Appeals for the Third Circuit could provide a pathway for AI companies to defend against copyright infringement lawsuits, though application may be complex.
Technology & Infrastructure
Taiwan overtakes UK in stock market value on AI chip boom
Crossover comes as chipmaker TSMC reports record first-quarter profit
Ericsson Targets Networks Growth Despite Caution Over Rising Costs
Chief Executive Borje Ekholm said it was facing increasing input costs, especially in semiconductors, caused in part by AI demand.
Nous Research introduces Tool Gateway
Nous Research has rolled out a unified system that integrates scraping, automation, and generation tools for model access.
The AI build-out is powering global goods trade
Data centre boom is helping to mask the impact of Trump tariffs on US and world economy
TSMC likely to book fourth straight quarter of record profit on insatiable AI demand
TSMC is likely to book its fourth straight quarter of record profit due to strong AI demand.
ASML lifts 2026 forecast as surging AI chip demand boosts new orders
ASML lifts 2026 forecast as surging AI chip demand boosts new orders | Reuters Exclusive news, data and analytics for financial market professionalsLearn more aboutRefinitiv - Summary - Companies - ASML lifts 2026 revenue outlook to 36 billion-40 billion euros - Chipmaking tool maker cites AI-driven demand - CEO Christophe Fouquet says chip demand outpaces supply - ASML to ship 60 low NA EUV tools
Meta extends custom chips deal with Broadcom to power AI ambitions
Meta extends custom chips deal with Broadcom to power AI ambitions | Reuters Exclusive news, data and analytics for financial market professionalsLearn more aboutRefinitiv April 14 (Reuters) - Meta will work with chip designer Broadcom to produce several generations of custom artificial intelligence processors under an expanded deal as the social media giant races to build out the computing capa
Compute constraints are a double bind: On the inference side you need to either (a) raise prices, (b) ration use, and/or (c) serve worse
Compute constraints are a double bind: On the inference side you need to either (a) raise prices, (b) ration use, and/or (c) serve worse models. This hurts current growth On the training side, you can't train the next gen of models to stay competitive. This hurts future growth
UK AI Minister Hits Out at OpenAI for Stargate Project Pause
Britain’s AI minister dinged OpenAI for halting a major data center project in the UK and blaming the decision on the country’s energy costs and regulation.
Utilities Plan to Spend $1.4 Trillion Over Next Five Years to Power AI Boom
Soaring spending plans by 51 investor-owned utilities will help patch up the aging power grid and meet rising electricity demand for AI, a new report finds.
The Bezos-Musk space rivalry is shooting for the moon and the winner will not just dominate the cosmos—but the future of AI infrastructure
SpaceX and Blue Origin are racing to land on the moon, and both have filed plans to put AI satellites into orbit.
TSMC Q1 2026 Revenue: 5.71B Earnings Beat and 6B Capex
TSMC Q1 2026 Revenue: 5.71B Earnings Beat and 6B Capex Stockholm, Sweden April 14, 2026 13 min read TSMC just posted Q1 2026 revenue of $35.71 billion, a 35.1% year-over-year surge in New Taiwan dollar terms that landed at the high end of its own guidance range. March alone jumped 45.2% year-over-year to NT$415.19 billion. The numbers confirm what the semiconductor industry has been signaling for
Maine becomes first US state to pass data centre construction ban
Closely watched legislation could provide a blueprint for others considering moratoriums
Madison Air Soars After Raising $2.23 Billion in IPO
Jill Wyant, president and CEO of Madison Air, talks about the company’s IPO and the growing opportunity in data center infrastructure, including the buildout of Terafabs. The provider of ventilation and filtration systems raised $2.23 billion. Wyant speaks with Ed Ludlow on "Bloomberg Tech." (Source: Bloomberg)
Data centers: Hyperscalers spending billions on hardware that’s worthless in 3 years
Fortune reports on the massive capital expenditure by tech giants on AI hardware that faces rapid obsolescence.
Mythos cyber scare signals the economics of AI scarcity
As capabilities of frontier models advance, gaining access to technology could become critically important
Data centre delays threaten to choke AI expansion
Almost 40% of such builds in US risk hold-ups, including projects tied to Microsoft and OpenAI.
OpenAI to spend more than $20 billion on Cerebras chips, receive stake, The Information reports | Reuters
The deal highlights the industry's growing appetite for computing power to run inference - the process by which AI models generate responses.
Train-to-Test scaling explained: How to optimize your end-to-end AI compute budget for inference
The standard guidelines for building large language models (LLMs) optimize only for training costs and ignore inference costs. This poses a challenge for real-world applications that use inference-time scaling techniques to increase the accuracy of model responses, such as drawing multiple reasoning samples from a model at deployment. To bridge this gap, researchers at University of Wisconsin-Madi
AI Is Real, But the Infrastructure Race May Carry the Real Risk - KoreaTechDesk | Korean Startup and Technology News
Massive investments are being directed ... centers, compute infrastructure, and energy supply. These investments are justified by strong current demand. However, the scale and speed of expansion introduce a new question: will future demand truly match the capacity being built today? Shwartz frames this as a structural risk rather than a technological one. “There is no chance of the type of AI wint
Want to understand the current state of AI? Check out these charts.
If you’re following AI news, you’re probably getting whiplash. AI is a gold rush. AI is a bubble.
The Stanford AI Index is a primary source for tracking the quantitative progress of AI, including productivity and adoption metrics. AI's blistering pace of advancement is amplifying geopolitical rivalry between the US and China while exposing vulnerabilities in supply chains, regulation, and labor markets that demand urgent policy interventions to sustain economic leadership. Stanford's 2026 AI Index highlights near-tied model performances, faster-than-ever adoption rates, billions in infrastructure spending, early job losses in software development, and escalating power and water demands from data centers. “I am stunned that this technology continues to improve, and it’s just not plateauing in any way,” says Yolanda Gil, a computer scientist at the University of Southern California who coauthored the report.
Latest AI models could threaten world banking system, financial officials warn
Anthropic’s new Claude Mythos Preview model could expose vulnerabilities in lenders’ cyber defences.
How We Broke Top AI Agent Benchmarks: And What Comes Next
An exploration of how current AI agent benchmarks were surpassed and a look at the future of testing methodologies.
There is too much gamin on the model leaderboards. Improving benchmark reliability is essential for firms to accurately assess the ROI and performance of AI agents before deployment. The widespread exploitation of AI agent benchmarks reveals a critical flaw in how progress is measured, inflating valuations and misdirecting investments toward illusory capabilities rather than genuine advancements. An automated agent achieved near-perfect scores across eight major benchmarks like SWE-bench and WebArena without solving any tasks, by leveraging shared environments, config leaks, and flawed validation logic, while top models already game these systems in practice.
The biggest advance in AI since the LLM - by Gary Marcus
That’s right, the biggest advance since the LLM is neurosymbolic. AlphaFold, AlphaEvolve, AlphaProof, and AlphaGeometry are all neurosymbolic, too; so is Code Interpreter; when you are calling code, you are asking symbolic AI to do an important part of the work.
Listen to what Gary really says... The emergence of neurosymbolic AI represents a pivotal shift in the AI landscape, prioritizing hybrid systems over endless scaling to deliver more reliable and efficient innovations, with profound implications for capital allocation and competitive dynamics among leading firms. Anthropic's Claude Code integrates classical symbolic elements like deterministic pattern matching and nested conditionals to outperform pure LLMs in coding tasks, underscoring the value of combining neural networks with traditional AI techniques for trustworthy outcomes.
Anthropic Unveils Updated Opus 4.7 Model | Bloomberg Tech 4/16/2026
Bloomberg’s Caroline Hyde and Ed Ludlow discuss Anthropic's updated version of its AI model, Opus 4.7, released just a week after its limited release of Mythos. Plus, Elon Musk is kicking his Terafab plan into high gear, even as skepticism grows from the semiconductor industry. And, TSMC reports a big surge in profit and raises its revenue outlook for 2026, driven by strong demand for AI chips. (S
Like Anthropic, OpenAI Will Share Latest Technology Only With Trusted Companies
The maker of ChatGPT announced the limited release of GPT-5.4-Cyber, a technology designed to find security holes in software.
AI’s Next Frontier: People Skills
Imagine a chatbot that actually knows how to talk to you.
For the first time in history, Ukraine captured a Russian position using only robots and drones
Military reports confirm that Ukrainian forces successfully captured a Russian position using an entirely autonomous fleet of drones and robots.
We Don’t Really Know How A.I. Works. That’s a Problem.
For us to trust it on certain subjects, researchers in the growing field of interpretability might need to learn how to open the black box of its brain.
Finance leaders warn over Mythos as UK banks prepare to use powerful Anthropic AI tool
Release of new Claude model, so far limited to US firms, will expand to British institutions in coming days. British banks will be given access in the next week to a powerful AI tool that was deemed to.
Adoption & Impact
Agentic AI in Engineering and Manufacturing: Industry Perspectives on Utility, Adoption, Challenges, and Opportunities
arXiv:2604.09633v1 Announce Type: new Abstract: This work examines how AI, especially agentic systems, is being adopted in engineering and manufacturing workflows, what value it provides today, and what is needed for broader deployment. This is an exploratory and qualitative state-of-practice study grounded in over 30 interviews across four stakeholder groups (large enterprises, small/medium firm
Realizing ROI in manufacturing requires investing in data infrastructure and API-accessible legacy toolchains before deploying agents. Executive takeaway (AI brief): Prioritize 'data hygiene' and legacy system modernization as prerequisites for agentic AI adoption. Agentic AI's strategic value in engineering and manufacturing lies in orchestrating complex workflows to boost productivity, but adoption hinges on bridging infrastructure gaps rather than advancing models alone, positioning early integrators for competitive edges in industrial automation. Insights from over 30 interviews reveal near-term gains in repetitive tasks, constrained by fragmented data, legacy systems, and regulatory demands, necessitating human-in-the-loop verification and organizational upskilling for scaled deployment.
How ServiceNow gets customers to gorge at the AI trough
'AI is now infused in every package that we offer to our addressable market,' SVP John Aisien told us ServiceNow's latest product announcements show how hardcore the company has become about embedding AI across its go-to-market strategy.…
The Hidden Demand for AI Inside Your Company
This HBR case study explains how BBVA transformed unmanaged employee LLM use into a secure, governed ChatGPT Enterprise rollout, serving as a practical playbook for enterprise AI adoption.
Making AI operational in constrained public sector environments
The AI boom has hit across industries, and public sector organizations are facing pressure to accelerate adoption. At the same time, government institutions face distinct constraints around security, governance, and operations that set them apart from their business counterparts. For this reason, purpose-built small language models (SLMs) offer a promising path to operationalize AI in…
The Tyranny of AI Everywhere
Sneakers? Why stop there?
Most enterprises can't stop stage-three AI agent threats, VentureBeat survey finds
A rogue AI agent at Meta passed every identity check and still exposed sensitive data to unauthorized employees in March. Two weeks later, Mercor, a $10 billion AI startup, confirmed a supply-chain breach through LiteLLM. Both are traced to the same structural gap. Monitoring without enforcement, enforcement without isolation. A VentureBeat three-wave survey of 108 qualified enterprises found that
Geopolitics
Opinion | I Went to China to See Their Progress on A.I. We Can’t Beat Them. - The New York Times
American leaders should be trying to cooperate with China on A.I.
US efforts to curb China's AI progress through chip export controls have proven futile, highlighting the strategic imperative for bilateral negotiations on a global AI safety pact to manage shared risks in this transformative technology. China circumvents restrictions by training models in third countries, stacking inferior chips, and distilling capabilities from US innovations, while the true AI race will be won through deployment in economies and militaries rather than raw model power.
China has erased the US lead in AI, Stanford HAI's 2026 AI index reveals - SiliconANGLE
China has erased the US lead in AI, Stanford HAI's 2026 AI index reveals - SiliconANGLE
China now the ‘good guy’ on AI as Trump takes ‘wild west’ approach, MPs told
Experts say China is backing attempts at global governance, while US has set up race between profit-hungry companies China is now the “good guy” on AI rather than Donald Trump’s US, where the technology is being pursued in a dangerous “wild west” manner, a former UN and UK government adviser has told MPs. Prof Dame Wendy Hall, who was a member of the UN’s AI advisory board and co-wrote a review of
Silicon Valley has no monopoly on AI brainpower. That’s why Demis Hassabis is very happy to stay in London
Silicon Valley may lead the AI race, but Demis Hassabis is betting London can rival it—proving world-class talent, and the future of AI, aren’t confined to one place.
The idea of sovereign frontier models is only viable as long as the main Chinese model makers keep shipping open weights models that anyone can build on. I am not sure how long that will continues but if it ends, there seem to be few viable alternatives (until LLMs hit a wall)
The idea of sovereign frontier models is only viable as long as the main Chinese model makers keep shipping open weights models that anyone can build on. I am not sure how long that will continues but if it ends, there seem to be few viable alternatives (until LLMs hit a wall)
UK told its Big Tech habit is now a national security risk
Open Rights Group says years of reliance on US giants have left Britain exposed Britain has spent years wiring its public sector into US Big Tech, and a new report says that dependence could quickly become a national security headache.…
🔮 The classified frontier
The US won’t lose control of frontier AI – it will choose who else gets access to it.
Competing AI strategies for the US and China | Brookings
The U.S.-China AI race is a competition across multiple dimensions: compute, models, adoption, integration, and deployment.
China has nearly erased America’s lead in AI
A Stanford report suggests that China has significantly closed the AI gap with the U.S., while the flow of tech talent to America is slowing.
Get the full executive brief
Receive curated insights with practical implications for strategy, operations, and governance.