Anthropic launches Fable 5.2 and Mythos 5.2 with lower prices and stronger vulnerability discovery
- Severity
- Info
- Sources
- 1
What happened
Anthropic released Claude Fable 5.2 and Mythos 5.2, which share a model but differ in safeguards; Mythos access is restricted to vetted cybersecurity and life sciences organisations. In Anthropic's internal CVE Rediscovery evaluation of 1,829 previously disclosed vulnerabilities, Mythos 5.2 scored 89.6% against 78.3% for Mythos 5.1, and it is also said to reproduce flaws and develop fixes better. Customers not eligible for zero data retention face 30-day retention for safety monitoring, and Enterprise Frontier Safeguards is in private beta. Defenders should confirm retention terms in writing before sensitive use and keep agent activity isolated, monitored and subject to human approval.
For security analysts · Written by AI from 1 source · Wrong?
What to do
- Confirm zero data retention arrangements in writing before using these models for sensitive workloads.
- Run AI agents in isolated environments, monitor their activity and require human approval for critical changes.
Written by AI from the sources, checked against them
Full story timeline · 1 update
Who reported · 1 outlet
Who and what is in it
Companies
Products