Anthropic blocks possible attempt to use AI to make biological weapons

Anthropic says it identified and disrupted attempts to use its AI models for “malicious activity” that could support biological weapons development. The finding comes from its latest threat intelligence report, released on Thursday, and it lands in the middle of a louder argument over what frontier AI can already do, and what it might soon do better.

The company says it blocked “malicious use” of Claude Haiku, Sonnet and Opus between December 2025 and August 2026. It also said there were five case studies involving actors using its models in ways that could support biological weapons development. Jacob Klein, Anthropic’s head of threat intelligence, told the New York Times that this was “an incredibly nuanced situation” rather than the cartoon version of someone openly trying to build a weapon.

Anthropic’s report goes well beyond biosecurity. It says Claude was used in cyber operations, influence campaigns, surveillance, scams, fraud and conventional weapons work. The company said it detected use tied to a Russia-based cyber espionage campaign, an Iranian propaganda institution, and groups including ShinyHunters, while also accusing Chinese AI firms of trying to replicate Claude’s capabilities. It said one hacking group consistent with Russia-based Midnight Blizzard used AI to help build malware that could spot security defenses and rewrite itself until it slipped past detection.

The company also says it saw six cases where Claude was used to develop software for conventional weapons, including firearms, missiles, armed drones and bombs, plus the systems that guide them. And it says its researchers found attempts ranging from fake dating apps and hotel WiFi scams to surveillance aimed at identifying dissidents. Anthropic says it has folded the findings back into its own safeguards and shared intelligence with authorities and industry partners where appropriate.

This report arrives as the political noise around AI gets louder, not quieter. Bernie Sanders wants a pause on advanced AI and a ban on artificial superintelligence. Donald Trump says the U.S. can’t afford to lose the race. Meanwhile, Anthropic’s own safety researcher has warned of a greater than 10% chance AI could kill all humans within the next decade. That is not exactly a calming public-relations backdrop.