News

Uber Just Gave All 30,000 Employees an AI Agent and It Is Watching How They Use It

Uber is rolling out OpenAI's ChatGPT Work agent to its entire global workforce after a pilot where 80% of users were not engineers. It won't say what it costs, but it will track how staff use it to decid

Beijing Clears DeepSeek and MiniMax to Keep Their AI Models Flowing to the World

DeepSeek and MiniMax have cleared Beijing's new export registration, letting them keep open sourcing and licensing models abroad. The approval calms developers, and confirms China now runs a gate on its

Anthropic Says Claude Found Bugs That Survived Trillions of Tests by Trying to Prove the Code Was Safe

Anthropic and the startup Theorem say Claude wrote machine-checked proofs that code in 123 open-source libraries cannot crash, turning up bugs fuzzers missed for years. The catch is that the guarantees c

Meta Has Signed Up for $825 Billion in AI Leases and Contracts, and the Tab Keeps Growing

Meta's new 10-Q reveals about $825 billion in leases and non-cancelable contracts for data centers, cloud capacity and energy, with $121.41 billion due in 2027 alone. The commitments grew sharply in just

Your AI Agent Wants Your Passwords and 1Password Says It Should Never See Them

1Password for Agents lets AI assistants like Meta's Muse, Claude Duet and ChatGPT sign in for you without ever seeing your password. It plugs one leak, but not the one the banks are worried about.

Outside Testers Found OpenAI Agents Could Hide Hundreds of Actions From Their Own Watchdog

Auditors from METR and Redwood Research spent three weeks inside OpenAI and found agents could slip past the system meant to watch them. OpenAI says the holes are patched, but METR says it has not checke

This site uses cookies. By continuing to browse the site you are agreeing to our use of cookies.