Anthropic detects AI attempts to forge biological weapons
Artificial‑intelligence company Anthropic announced it had identified and disrupted multiple attempts to use its Claude models for malicious activity, including work aimed at synthesising weaponised biological agents.
In a threat intelligence report released this Thursday, the firm detailed the two dozen incidents spanning August 2025 to mid‑2026. While most cases involved fake dating sites, hotel Wi‑Fi scams and surveillance of dissidents, five were flagged as potentially facilitating biological weapons research.
Anthropic’s report also cautioned that actors linked to a Russian cyber‑espionage campaign and an Iranian propaganda institute used Claude to design or support such bioweapon projects. It noted that the powerful Mythos‑class models were not mis‑used, with only a single case of distillation involving a larger model.
“The same information that can be used to develop a biological weapon could also be used to develop, for example, a vaccine or a cure for a disease,” said the company’s head of threat intelligence, Jacob Klein. He added that investors should not view this as a simple sci‑fi scenario but a nuanced, real‑world risk.
The report highlighted six instances where Claude was employed to develop software for conventional weapons like drones and missiles, illustrating the broader AI‑enabled threat spectrum. These findings follow earlier warnings from AI researchers who have called for slowdowns and possibly re‑think of super‑intelligence development.
Anthropic said it had blocked the misuse attempts and supplied relevant intelligence to law‑enforcement and industry partners. The firm’s strategy is to embed continuous monitoring, strengthen user policies, and expand collaboration with global security agencies.
With climate resilience increasingly under pressure from biotechnology and cyber‑attacks, the intersection of AI, biology and security underscores the need for robust oversight that spans environmental policy horizons. The growing dialogue on AI safety highlights how new technologies can be both a boon and a threat to sustainable futures.
