Anthropic says AI misuse cases include attempts to support biological weapons development, cyber operations, surveillance, scams and fraud.
Anthropic says it has identified and disrupted attempts to misuse its artificial intelligence models in ways that could support the development of biological weapons.
The disclosure was contained in the company’s latest threat intelligence report, published Thursday, detailing cases of suspected misuse of its Claude AI models between December 2025 and August 2026.
Anthropic said it identified five cases involving actors using its models in ways that could support biological weapons development. The company described biological misuse as “one of the most serious risks of frontier AI models,” warning that without appropriate safeguards, such capabilities could have catastrophic consequences.
“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,” Anthropic said.
Jacob Klein, Anthropic’s head of threat intelligence, told the New York Times that the issue was “an incredibly nuanced situation,” noting that malicious intent is not always obvious.
“You are not seeing someone in a comic book kind of way say, ‘Hey, I want to build a biological weapon to kill everybody,’” he said.
The report also identified six cases in which Claude was allegedly used to develop software for conventional weapons, including firearms, missiles, armed drones, bombs and other munitions, as well as systems used for targeting and control.
Anthropic said its models had also been misused for cyber operations, surveillance, scams, fraud and influence campaigns. Cases identified over the past eight months included fake dating applications, hotel Wi-Fi scams and surveillance systems designed to identify dissidents.
The company said Claude had been used by actors linked to a Russia-based cyber espionage campaign and an Iranian propaganda institution. It also named hacking group ShinyHunters and China-based laboratories among those associated with misuse of its technology.
According to the report, a hacking group whose activities are consistent with Russia-based Midnight Blizzard allegedly used AI to develop a system capable of detecting when its malware had been identified by security systems and rewriting the code to evade detection.
Anthropic also accused Chinese AI companies of attempting to replicate Claude’s capabilities.
The company said it had incorporated the findings into its security processes “to better prevent, detect, and disrupt these activities in the future” and had shared relevant intelligence with authorities and industry partners.
The report comes amid growing concerns within the technology industry about the potential risks of increasingly powerful AI systems.
Anthropic said none of the reported misuse cases involved its Claude Fable or powerful Mythos-class models, apart from one case involving distillation, a process in which smaller AI models are trained using larger models.
The company’s warning follows separate concerns raised by AI researchers and industry figures over the pace of development.
OpenAI chief scientist Jakub Pachocki recently called for “voluntary slowdowns” in AI development until stronger safeguards are established, saying he was concerned about the consequences of a continued rapid rise in machine intelligence.
In the United States, Democratic Senator Bernie Sanders has called for a temporary pause on advanced AI development and introduced legislation seeking to ban artificial superintelligence.
President Donald Trump, however, has rejected calls to slow AI development, saying Thursday that he was concerned the US could be placed “in a very bad position” if it failed to win the AI race.
Anthropic’s report is the company’s first threat intelligence report of the year and adds to mounting warnings over the potential misuse of increasingly capable AI systems.
