Retrieving the latest stories.
The most consequential developments in technology and the world, selected against strict editorial criteria. Each is explained in twenty seconds.
Why it mattersGoogle is flipping K-12 defaults so Gemini and Search AI features are off for student accounts until administrators opt in, while Search starts nudging students away from direct answers, though the effect on learning rests on Google's own research.
Why it mattersGoogle will switch AI off by default for K-12 students from next year, but because AI is built into Search, schools that restrict AI lose Search entirely, which shows how much control Google has over classroom technology.
Why it mattersMeta's automated systems mistook a coalition's Claude-assisted social media monitoring for bot traffic and locked No Kings admins out for four hours during an early-voting push, prompting 14 attorneys general and lawmakers to question how AI agents and platfo…
Why it mattersThe EPA ended its 20-year effort to estimate factory-farm air emissions, closing the 2005 consent agreement and leaving a pollution source linked to about 12,700 US deaths a year with no federal measurement method, though the practical enforcement effect is s…
Why it mattersFOI responses show 29 of 137 NHS trusts credited with Palantir benefits say they don't use the platform, undermining the evidence behind a £330M contract decision due in December, though the story is UK-specific and partly campaign-sourced.
Why it mattersDaily security brief.
Why it matters43B Binance theft shows stablecoin issuers froze 44% of the loot (versus about 3% after Bybit) while North Korea's automated laundering of the rest will likely take months, which tells readers how much real leverage freezes and cash-out choke points give agai…
Why it mattersThe government's own AI pilot found hundreds of flaws, and the NCSC's pessimistic scenario names China as a likely supplier of AI attack tools.
Why it mattersMemory cost pressure is expected to keep squeezing margins, and Apple has already raised prices to offset it.
Why it mattersIn a revised S-1A, Anthropic says it may 'imminently' meet its own definition of dramatically AI-accelerated progress (recursive self-improvement), admits it has missed such accelerations before, and is weighing policy changes to keep humans in control.