Skip to content

Anthropic

How model auto-switching works in Claude Fable 5

Written by an AI model from the linked sources and checked against them by a second model. How we work

Image: support.claude.com

Anthropic has detailed the fallback mechanism from Fable 5 to Opus 4.8. Every request to Fable 5 undergoes automated safety checks across four sensitive domains, primarily cybersecurity and biology. If a trigger occurs, the request is rerouted to Opus 4.8 within the same conversation.

The checks analyze more than just the latest message. They scan everything the model reads, including memory, connector data, web search results, and files. Content the user did not explicitly input can still trigger a block.

Once the switch occurs, the session stays on Opus for the remainder of the dialogue, though users can manually revert to Fable 5. Auto-switching is enabled by default in Claude, Cowork, Code, Design, and Claude for Microsoft 365, but it can be disabled in settings. If a request is blocked even on Opus, users can submit a request to the Cyber Verification Program for sensitive cyber-related tasks.

Anthropic

Anthropic to develop its own pharmaceuticals

Anthropic is moving into drug development. After launching Claude Science for researchers, the company is shifting from selling AI tools to pharmaceutical firms to developing its own proprietary medicines. Specifics on their therapeutic focus remain undisclosed.

No comments yet. Be the first.

Join the conversation

Sign in to leave a comment. It appears after moderation.