Artificial intelligence company Anthropic has identified and disrupted attempts to use its AI model for malicious activities that could support the development of biological weapons. The findings were detailed in the firm's recent threat intelligence report, underscoring a growing wave of warnings from AI researchers and industry insiders about the potential risks posed by advanced artificial intelligence.

These warnings have intensified calls for stricter oversight and regulation of AI development. Notably, US Senator Bernie Sanders has advocated for a pause in advanced AI development and a ban on artificial superintelligence. In contrast, President Donald Trump has expressed a different concern, stating on Thursday that "if we don't win AI, we're going to be put in a very bad position."

The report highlights that Anthropic's AI model, Claude, was also reportedly used by actors linked to a Russia-based cyber espionage campaign and an Iranian propaganda institution. Over the past eight months, detected cases of concern ranged from the creation of fake dating apps and hotel Wi-Fi scams to surveillance tools designed to identify dissidents. Anthropic also accused Chinese AI firms of attempting to replicate Claude's capabilities.

Anthropic's lengthy report lists various instances where suspected state-sponsored groups, criminals, spyware vendors, state propaganda institutions, and politically motivated individuals misused its technology. The company stated that malicious use of its Claude Haiku, Sonnet, and Opus models was disrupted between December 2025 and August 2026. The report excluded its Claude Fable and Mythos-class models, except for one instance of "distillation," a process used to train smaller AI models from larger ones.

Beyond detecting the use of its models for cyber and influence operations, surveillance, scams, fraud, and weapon development, Anthropic reported blocking scientists whose AI use could support biological weapons development. The report specifically highlighted five case studies where actors used its models in ways that could aid in biological weapons development, labeling biological misuse as one of the most serious risks associated with frontier AI models.

Anthropic emphasized that without proper safeguards, such capabilities could have "catastrophic consequences." The company noted the dual-use nature of the technology, stating that the same information used to develop a biological weapon could also be used to develop a vaccine or cure for a disease.

Jacob Klein, head of threat intelligence at Anthropic, described the situation as "incredibly nuanced" to The New York Times. He clarified that these were not overt requests to build weapons, but rather subtle utilizations of the AI's capabilities. The report also documented six cases where Claude was employed to develop software for conventional weapons, including firearms, missiles, armed drones, bombs, and other munitions.

The report's findings add to a growing body of evidence suggesting the need for robust safety protocols and international cooperation in managing the development and deployment of powerful AI systems. The incidents raise critical questions about the responsibility of AI developers and the effectiveness of current safety measures in preventing misuse.