Meta will log every mouse movement, click, and keystroke from its US employees to train AI agents. The same workforce facing 10% layoffs next month is generating the data that could automate their jobs.
Daniel Moreno-Gama brought a Molotov cocktail and a document vowing to kill AI executives to Sam Altman's doorstep. The most prominent act of anti-AI violence on record is heading to court.
The EU's 2027 battery mandate doesn't mean snap-off back panels are coming back. But the days of aggressively glued-together phones are numbered — and the ripple effects could reshape global hardware design.
Four months after reports that Apple's top chip designer was considering leaving, Johny Srouji has been promoted to run all of the company's hardware. The move locks down Apple's most important technical advantage before John Ternus takes over as CEO.
Tim Cook grew Apple from $350 billion to $4 trillion. His replacement is a mechanical engineer who has spent 25 years building hardware and has never run a company. The markets will decide whether product instincts beat operational discipline.
Tesla's robotaxi expansion to Dallas and Houston lasted roughly 48 hours before availability flatlined. Tracking data suggests the company deployed one car per city — three days before its Q1 earnings call.
French prosecutors have summoned Elon Musk over Grok's generation of millions of sexualised deepfakes on his own platform. Whether the world's richest man shows up is the immediate question.
The NSA is running Anthropic's Mythos model — and expanding it — despite a formal Pentagon blacklist. Two days earlier, Trump said he had "no idea" his own chief of staff met with Anthropic's CEO.
Blue Origin just became the second company ever to land a reused orbital-class booster. The commercial satellite it was carrying may be in the wrong orbit.
Honor's Lightning robot finished Beijing's half-marathon in 50:26 — seven minutes faster than any human has ever run the distance. Twelve months ago, most of these machines couldn't complete the course.
Three more executives walked out Friday as OpenAI killed the projects they led. The talent is flowing straight to Anthropic, Meta, and Google — and the company's $852 billion valuation has not yet priced that in.
A Starlink service disruption knocked out US military drone tests this month. The real story isn't the outage — it's that the Pentagon has no backup.
A mainstream AI model — not the restricted Mythos — built a full Chrome exploit chain for the price of a laptop. The patch gap in millions of desktop apps just became a whole lot more exploitable.
A design flaw baked into the Model Context Protocol lets attackers execute arbitrary commands on up to 200,000 servers. Anthropic calls it expected behavior.
The AI model too dangerous for public release is about to become standard infrastructure across the US federal government — by order of the same administration that spent February trying to destroy the company that built it.
OpenAI will pay Cerebras over $20 billion and take an equity stake of up to 10%, according to The Information. The deal comes as Cerebras prepares to refile for an IPO that could value it north of $23 billion.
A fraud defendant's conversations with Claude are now prosecutorial evidence. More than a dozen major US law firms have issued urgent warnings: your AI chat logs have no legal protection whatsoever.
A Swedish thermal plant, a Norwegian dam, a Polish power grid. Russian hackers are no longer just reading systems — they're trying to break them.
Meta tried to satisfy EU antitrust regulators by charging rivals for access to WhatsApp. The EU's response: charging what amounts to a toll is no different from locking the gate.
Snap eliminated 1,000 jobs and said the quiet part out loud: AI made the work unnecessary. More than 65% of its new code is now machine-generated.
lop News
Jonny V. Nuovo