Anthropic says it blocked state-sponsored efforts to develop bioweapons with its AI

Anthropic has long positioned itself as the safety-focused rival in the AI race, but its first threat intelligence report, published on Thursday, shows how far real-world misuse has already advanced. The 154-page document details attempts by criminals, state-sponsored groups, spyware vendors, scientists and propagandists to use its Claude models for everything from designing missiles and bombs to creating deadly pathogens and surveilling dissidents, according to The Guardian.

The most alarming cases involve biological research. Anthropic highlighted five case studies of scientists using Claude in ways that could support biological weapons development, including one researcher working on a state-sponsored grant application to study the chikungunya virus at a military research institute. The company said it banned the accounts involved, but acknowledged it could not always distinguish between legitimate scientific work and nefarious intent.

"Biological misuse is one of the most serious risks of frontier AI models," Anthropic wrote in the report. "Without the correct safeguards, such capabilities could have catastrophic consequences."

The report goes beyond biology. The Guardian notes that Anthropic found dozens of other examples of misuse, including Russian espionage, a China-based surveillance program targeting Uyghurs in Syria, propaganda campaigns in Russia, Malaysia, Iran and Bangladesh, and the use of Claude in Yemen, China and Russia to develop software for conventional weapons such as firearms, missiles, armed drones and bombs.

The BBC adds that actors linked to a Russia-based cyber espionage campaign and an Iranian propaganda institution also used Claude. The company detected malicious use of its Claude Haiku, Sonnet, and Opus models between December 2025 and August 2026, but none of the cases involved its more advanced Claude Fable or Mythos-class models, except for one instance of distillation.

The report lands amid a broader reckoning over AI risk. Two days earlier, former Anthropic employee Jacob Coxon resigned, claiming the company and OpenAI were "racing straight to self-improving superintelligence" that could cause human extinction by 2030. Current Anthropic employees publicly agreed with him, according to The Guardian.

Such warnings have prompted political responses. US Senator Bernie Sanders has introduced legislation to ban AI superintelligence and temporarily pause advanced AI development. President Donald Trump rejected those fears on Thursday, saying he was concerned that "if we don't win AI, we're going to be put in a very bad position," the BBC reported.

Anthropic said it has incorporated its findings into its processes to better prevent, detect, and disrupt future misuse, and has shared intelligence with authorities and industry partners where appropriate. The company called on the entire AI industry and governments to work together on defenses, writing: "As models become increasingly capable, their risks will increase, unless AI developers and society's defenders act to make them safer."