Anthropic’s confirmation of a physical wet lab in the San Francisco Bay Area for biology experiments has emerged as the key driver of trader sentiment around potential lab-leak risks. The facility supports Claude-powered protein design, robotic automation of experiments, and work on rare-disease targets, coinciding with the company’s September 2026 threat-intelligence report detailing blocked attempts to use its models for gain-of-function research on viruses such as chikungunya and avian influenza. This expansion into real biological material handling, while Anthropic simultaneously publishes metrics showing its models now leading 26% of internal R&D and urges industry-wide slowdowns, underscores the tension between capability gains and biosafety controls. Traders are watching for upcoming Life Sciences Verification Program grants, independent evaluations with partners like Accenture, and any further disclosures on lab operations or model releases that could alter perceived incident probabilities.