Anthropic introduces 30-day data retention for Mythos and Fable models

Starting June 9, 2026, Anthropic is updating its data handling policy for its most powerful models. Requests and responses for Mythos and Fable class models will be stored for 30 days for security purposes.
The change affects organizations that previously used the zero-data retention (ZDR) mode, including users on Claude Enterprise, AWS Bedrock, Google Cloud Agent Platform, and Microsoft Foundry. Consumer users on standard plans are unaffected.
Anthropic cites the need to detect abuse patterns like repeated jailbreak attempts. Threat analysis requires reviewing sequences of requests, not individual messages. Anthropic staff access to data is restricted and all actions are logged.
Up next
Anthropic updates safety policy amid valuation surge
Anthropic drops binding safety commitment. The updated policy replaces a binding pause on model development with internal targets. This change comes as Anthropic’s valuation surged to $380B by Feb 2026, ahead of a planned IPO.
Related stories
- Claude Code auto-deletes session history after 30 days
- Claude Security moves to continuous scans on Enterprise
- Anthropic adds offensive tiers for verified pentesters
- Mythos 5 lands in Claude Security for enterprise beta
- Anthropic plans to loosen data retention on top models
- Rubrik found more bugs than its engineers could fix
Comments
No comments yet. Be the first.
Join the conversation
Sign in to leave a comment. It appears after moderation.

