Anthropic Biosecurity Watch & Dual-Use AI Research Inspector
Interactive simulation of frontier AI safety guardrails, biological threat vector analysis, and automated plot halting mechanisms inspired by the NYT report on intercepted biological research attempts.
AI Safety Sandbox Controlsv2.6 Inspector
75%
Triggers automated plot halting and institutional flagging when risk exceeds threshold.