In May, Anthropic implemented “AI Safety Level 3” protection alongside the launch of its new Claude Opus 4 model. The ...

A new feature with Claude Opus 4 and 4.1 lets it end conversations with users with "persistently harmful or abusive ...

Discover how Anthropic's Claude Code processes 1M tokens, boosts productivity, and transforms coding and team workflows. Claude AI workplace ...

Testing has shown that the chatbot shows a “pattern of apparent distress” when it is being asked to generate harmful content ...

Claude won't stick around for toxic convos. Anthropic says its AI can now end extreme chats when users push too far.

It will only activate in "rare, extreme cases" when users repeatedly push the AI toward harmful or abusive topics.

Anthropic’s Claude Code now features continuous AI security reviews, spotting vulnerabilities in real time to keep unsafe ...

By empowering Claude to exit abusive conversations, Anthropic is contributing to ongoing debates about AI safety, ethics, and ...

Dario Amodei said he believes Anthropic employees are largely staying because of "true belief in the mission and belief in ...

The new context window is available today within the Anthropic API for certain customers — like those with Tier 4 and custom ...

Anthropic launches automated AI security tools for Claude Code that scan code for vulnerabilities and suggest fixes, ...

Can exposing AI to “evil” make it safer? Anthropic’s preventative steering with persona vectors explores controlled risks to ...