
EU Regulators Back Anthropic’s Staged Claude Mythos Release Over Cyber Risk Concerns
EU regulators welcomed Anthropic’s decision to slow the rollout of Claude Mythos after findings that the model could outperform humans at identifying and exploiting software vulnerabilities; a group of 12 cybersecurity firms and 40 other unnamed organizations have access to test defenses, and the EU is in dialogue with Anthropic to ensure the model has adequate cybersecurity protections under EU AI rules.


