Anthropic’s new Bay Area wet lab, focused on BSL-1 and BSL-2 molecular biology work with human researchers conducting all physical experiments, underpins the low implied probabilities of 3% by year-end 2026 and 9% by mid-2027 on a pathogen escape. The facility supports AI-assisted discovery, including a recent CRISPR-like enzyme system identified via Claude agents, but avoids high-risk human-infecting pathogens and gain-of-function research on select agents. Traders price in robust containment, Anthropic’s safety emphasis, and the absence of prior lab-leak precedents at AI companies, though expanded capabilities in synthetic biology or any regulatory shift on dual-use AI research could introduce tail risks. No confirmed incidents have surfaced in credible reporting since the lab’s public confirmation.