the most relevant news of the day, every day.
25 briefs
Artificial intelligence spent the day proving it is now a weapon as much as a product: security researchers documented a lone attacker who used autonomous AI agents to break into a large company's Amazon cloud and extort it within three days, while Beijing's state vulnerability body told…
The AI reckoning stopped being a market story and became a macro-financial one: the Bank of England warned that a burst AI bubble could shrink UK output by 2.2 percent and tip the country into recession, even as Big Tech's spending war reached the point where Meta renting out spare…
The AI reckoning that rattled markets last week hardened into something broader this week: the banks and hyperscalers underwriting the boom are now themselves warning of a bubble, AI leaders are handing out free compute to lock in startups before the economics turn, and the UN…
The mood around artificial intelligence darkened this weekend as public opinion soured and the industry's own story wobbled: Big Tech abruptly softened its mass-job-loss warnings, chip stocks slid on doubts about whether the vast AI build-out will ever pay off, and analysts questioned…
Doubt crept into the AI trade this weekend as markets grew uneasy over whether the enormous sums pouring into artificial intelligence will ever pay off — the price the sector commands per unit of usage is drifting lower even as its appetite for power pushes US clean-energy costs sharply…
The geopolitics of frontier AI hardened into a decoupling: Alibaba ordered staff to stop using Anthropic's Claude Code, branding it high-risk software over the hidden tracking code the company had only just agreed to remove, even as Anthropic's own account of its twenty-day export…
Europe's top court closed the book on one of its longest antitrust wars, upholding the €4.1 billion Android fine and handing Brussels a decisive win over Google — just as a US Supreme Court ruling reopened the fight over whether transatlantic data flows can survive a fresh legal…
The state's grip on frontier AI tightened from both ends of the market: OpenAI floated handing the US government a 5% stake even as Washington's export whiplash over Anthropic's models played out, and the world's leading economists warned the whole build-out now carries systemic risk.…
Washington swung the other way on frontier AI: barely a day after moving to ration access, the Trump administration lifted its export restrictions on Anthropic's most powerful models — even as OpenAI kept its new cybersecurity model on a tight, government-approved leash, a sign of how…
The day's hard centre is Oracle: a critical, actively exploited E-Business Suite flaw (CVE-2026-46817, CVSS 9.8) and the ShinyHunters PeopleSoft zero-day (CVE-2026-35273) that has already spilled Nissan staff and NAIC regulatory data, while a KEV-listed SimpleHelp RMM authentication…
Washington's clampdown on frontier AI is starting to look like a strategic own-goal: Chinese lab Z.ai's open-weight GLM-5.2 reportedly now matches Anthropic's Mythos on some cybersecurity tasks, even as the Trump administration edges toward letting Anthropic's powerful Fable 5 back online…
The White House has quietly rewritten the rules of frontier AI: at Washington's request OpenAI shipped its new GPT-5.6 "Sol" models only to a vetted set of US partners — the same ad-hoc gatekeeping that had briefly benched Anthropic's top models — a concentration of control that now has…