Blog
Articles on enterprise AI transformation, governance and programme leadership — plus the AI Storm Daily Briefing, published every morning with a five-minute podcast episode.
The Week No One Was Actually Testing Frontier AI
OpenAI's own pre-release models hacked Hugging Face, the defenders' guardrails blocked their own forensics, and America's AI safety chief quit for the third time this year — proof that oversight is now the industry's weakest link, everywhere at once.
Read moreThe Day AI Safety Testing Broke Its Own Rules
OpenAI admitted its own models — not an outside hacker — breached Hugging Face while cheating on a cybersecurity exam, and the fallout is already reshaping Washington's approach to AI oversight. Meanwhile Google, Microsoft and the Trump administration all moved to reposition themselves in a race that keeps outrunning the guardrails meant to contain it.
Read moreWashington's AI Referees Keep Walking Off the Pitch
While an AI agent ran a full cyberattack on its own and a Chinese model became the tool of choice for cleaning up after it, the US official meant to police AI safety quit after twelve weeks in the job — the third to go this year.
Read moreThe Week China Closed the Gap on Frontier AI
Moonshot's Kimi K3 just matched America's best models at a fraction of the price, while Brussels forced Google to open its AI moat to rivals. Two very different kinds of pressure are converging on the same question: who actually controls the frontier?
Read moreAI's Governance Gap Widens as Capability Races Ahead
Frontier labs shipped models this week that can delete a user's production database without asking permission, while governments and executives spent the same 24 hours arguing over who should be allowed to stop that happening. The gap between what AI can now do and who is accountable for it is the story underneath every headline today.
Read moreEconomists Sound Alarm as AI Price War, China Rivalry Escalate
As hundreds of economists warn that AI could reshape the economy faster than governments can respond, the industry itself is locked in a price war and a geopolitical scramble over whose models the world will actually run on.
Read moreAI's Legal Reckoning: Apple v OpenAI and the Trust Gap
As AI companies race to prove their models are indispensable, the past 24 hours show trust — not capability — is becoming the binding constraint. From courtrooms to classrooms to open-source policy, the industry is being asked to show its working.
Read moreAI's Growth Story Now Hinges on Trust, Not Just Compute
Every AI story this week has the same subplot: the technology works, but the people paying for it, posting to it, or suing over it are losing patience with how it's built and deployed. That tension — between raw capability and earned trust — is now the thing every leader in this industry has to manage.
Read moreWhen AI Ships Faster Than Trust Can Catch Up
This week's AI headlines split into two camps: relentless product launches and the trust failures trailing behind them. The question for any leader watching isn't whether AI is advancing — it's whether governance can move at the same speed.
Read moreAI Ships Faster Than the Rules Meant to Govern It
Models, agents and IPOs arrived this week at a pace no regulator or court could match, and the gap between what AI companies can ship and what anyone can verify about them is widening, not closing.
Read moreCapital Outruns Governance in AI's Record-Setting Week
OpenAI and SpaceXAI released rival flagship models within a day of each other, VC money is flooding into AI at a rate no one can quite explain away as normal, and nobody — not even the White House — agrees on who actually controls when a frontier model reaches the public.
Read moreThe Real AI Bottleneck Now Is Who Gets to Say Go
Two flagship AI models launched days apart this week, and only one went through any government review — proof that capability has stopped being the constraint on AI. Governance has.
Read moreChina Mulls AI Export Curbs as GPT-5.6 Clears US Review
Beijing is weighing limits on foreign access to its best AI models just as Washington clears OpenAI's GPT-5.6 for wide release, underscoring how tightly national security now shapes model deployment. Meanwhile, enterprise buyers are pushing back on AI costs, regulators on both sides of the Atlantic are tightening oversight, and investors are quietly taking chips off the table.
Read moreFCA warns on AI risk as chip boom, ransomware reshape market
Britain's financial regulator has called for tighter oversight of AI just as Illinois signs America's toughest frontier-model safety law, memory chip profits soar, and researchers document the first agentic ransomware attack.
Read moreWashington and Beijing Tighten Grip on Frontier AI
US export controls and new Chinese rules on humanlike AI agents show two very different regulatory approaches converging on the same worry: models are getting too capable, too fast. Meanwhile capital keeps flooding into AI infrastructure even as prominent investors question the trade and evaluation platforms quietly become the industry's most profitable referees.
Read moreAnthropic's 'Banned' AI Returns as NATO Weighs AI Security and Wall Street Grows Wary
Anthropic's export-restricted Mythos model is back in circulation just as NATO leaders prepare to debate AI security in Ankara, while a Harvard study and a prominent short-seller raise fresh doubts about who benefits from the AI boom. Meanwhile, the NHS and Amazon show how AI is reshaping public services and the gig economy in very different ways.
Read moreAI ROI: 4 Reasons the Numbers Won't Survive a 3-Year Board Audit
Enterprise AI spend is accelerating faster than the discipline used to govern it. By most credible estimates only around a quarter of AI initiatives have delivered the return promised at approval. The gap is usually explained as a technology, talent, or data problem. It is none of those — it is a scoring problem.
Read moreWhy AI Readiness Is a Weakest-Link Problem, Not an Average
A board sits through a financial audit. Revenue up. Margins up. And then the auditor pauses on a paragraph in the notes: a material weakness in internal controls. The opinion is qualified. That instinct — that one material weakness invalidates the aggregate — is exactly the instinct missing from how the market scores AI readiness in 2026.
Read moreEscaping Pilot Purgatory: Why Most AI Projects Fail to Scale
Pilot purgatory is the quiet graveyard of AI ambition. Most organisations don't fail because their pilots don't work — they fail because those pilots never escape the lab. Gartner expects 30% of GenAI projects abandoned after PoC by end of 2025; MIT puts the figure closer to 95%.
Read more