AI firms as first‑line biosecurity

Anthropic says it detected and disrupted several attempts this year to use its Claude model for biological research that could help engineer pathogens, and also flagged use for conventional weapons design and state propaganda. Those disclosures show that private AI companies are already acting as gatekeepers — deciding when to block queries, investigate actors, and inform authorities — despite limited public oversight or consistent standards. — If private firms are de facto biosecurity monitors, we need public debate about transparency, liability, standards, and how companies coordinate with governments to manage existential risks.