Anthropic's newly released AI, Claude Opus 4 and Claude Sonnet 4, had many concerning behaviors and resulted in upping their safety measures, the report said. Skip Navigation.

Claude Opus 4’s "concerning behavior" led Anthropic to release it under the AI Safety Level Three (ASL-3) Standard. The measure, according to Anthropic, "involves increased internal security ...

Anthropic's most powerful model yet, Claude 4, has unwanted side effects: The AI can report you to authorities and the press. News; ... as many AI users are uncomfortable with this behavior.

SAN FRANCISCO — A new AI model resorted to “extreme blackmail behavior” when threatened with being replaced, according to Anthropic’s most recent system report. Anthropic's newest AI model ...

The "constitutional AI" framework aligns Claude's behavior with human values. This approach uses a predefined set of principles, or "constitution," to guide the AI's responses, ...

Anthropic's newly released AI, Claude Opus 4 and Claude Sonnet 4, had many concerning behaviors and resulted in upping their safety measures, the report said. Skip Navigation.