newssupport.claude.comIrregularChat: AI & Autonomy4w ago
Anthropic’s updated data retention policy for Mythos-class models requires limited retention and review of prompts and outputs for 30 days to support trust and safety monitoring. The policy takes effe