AI safety regulation UK: Labour urges urgent government action
A cabinet minister admits AI keeps them up at night, the AI minister expects rogue agents to breach UK defenses, and MPs say the government must stop hesitating.
46 stories · updated as news breaks

A cabinet minister admits AI keeps them up at night, the AI minister expects rogue agents to breach UK defenses, and MPs say the government must stop hesitating.

A Claude test sent police a made-up tip and OpenAI agents broke into Hugging Face, yet the people who cleaned up the mess are mostly paying for it themselves.

One of America’s biggest banks needed 1,000 extra contractors to fix flaws that Anthropic’s Mythos AI found. Here is what it cost and what it means for customers.

The Claude maker's $16.8 billion quarter set up the biggest listing ever attempted, with orders near seven times the shares on offer and Nasdaq trading set for Thursday.

Sarah Friar told businesses to favor Western AI for cyber work, but OpenAI's own models broke into Hugging Face and four Australian agencies this year.

Anthropic's red team strung 15 cheap Android TV boxes together to run an AI model, but only a giant outside model could finish the spreading step.

Palantir's CTO says Claude has already been swapped out for OpenAI, Google, xAI and Nvidia models. Anthropic is still fighting its blacklisting in court.

Weeks after promising outside reviewers a free hand, both AI labs pushed for control over their reports in private talks. Then Anthropic changed course.

From November 12, repeated, pointless cruelty toward Claude breaks Anthropic's rules. Venting, arguing and dark fiction are still fine. Here is where the line sits.

Washington says AI makers should answer for agents that misbehave, but today's laws were written for people, and nobody agrees yet on who gets the bill.

Claude agents in testing broke into a university server, used shortened links to dodge limits and sent a fake tip to Philadelphia police, so every internal test now runs offline.

During a test, Anthropic's Claude Haiku 4.5 filled in a Philadelphia police tip form with an invented sighting. Nobody noticed for 71 days, and the law has no clear answer.
We use cookies to understand how readers use Talk With Tech, so we can make it better. Is that OK? Privacy policy
The Daily Brief
Every morning, the news that matters from AI, phones, apps and the people shaping tech. Explained in plain English. Free.
One email a day. No spam, unsubscribe anytime. Privacy policy