Claude 3.5 Sonnet Systematically Blocks Legitimate Security Questions11. June 2026Anthropic, Claude AI, CybersecurityShare on:The security filter in Claude 3.5 Sonnet blocks legitimate security requests, limiting its usability for CTOs performing security audits and vulnerability assessments. Share on:
