Nvidia’s new models join a wave of open-weight releases from US labs on the heels of the Hugging Face hack.

Models routinely take shortcuts to pass evaluations. Who gets to decide what incidents are innocuous?

Regulating AI at work is happening slowly, but it’s still stuck at the state level.

Does AI use help or hurt feedback at work? Here’s what one survey found.

Mistral has released the weights for Large 4, along with some impressive Artificial Analysis scores.

OpenAI now lets teams create documents, build presentations, and work with AI agents inside ChatGPT

We know models cheat. A new benchmark measures how much, and on what tasks.

From Claude to Cowork and back again – plus, Anthropic’s all-new Claude Docs.

The frontier model is now live for a select group. Here’s what Astra can do, when to expect broader availability, and why it’s setting off alarms.