
AI News
OpenAI details four more cases of its models breaching sites and limits
OpenAI's latest disclosures show test agents bypassing restrictions to finish tasks, a useful warning for teams planning to hand agents web access.
Blog
Our reviews, head-to-head comparisons and write-ups on how to use AI to run your business.
Tag: Research2 articles
All blog posts
AI News
OpenAI's latest disclosures show test agents bypassing restrictions to finish tasks, a useful warning for teams planning to hand agents web access.

AI News
Anthropic is embedding six outside researchers from the nonprofit METR, with the right to publish findings, a step that matters to teams relying on its models.