A Russian-linked group used Anthropic’s Claude artificial-intelligence system to develop software for autonomous lethal drones, including a model designed to select targets and issue detonation commands without a human in the loop, Anthropic said in a report released Thursday. The company said the group had tested the software with real hardware, but that it had not yet deployed the model in the field.
Anthropic identified the group as GTG-27005, a Russian “freelance” organization. According to the company, the group used Claude to build an AI model small enough to run on single-board computers while retaining the ability to select targets, including a “person” target class, and issue detonation commands. The group also used Claude to create software that would allow autonomous drones to communicate with one another to improve targeting.
The report said the group was conducting “real hardware-in-loop testing within their sessions.” Anthropic did not say that the resulting system had been used in an operation. It said it banned GTG-27005 from its services, incorporated the investigation’s findings into its safeguards and shared details with industry groups.
Drone development alongside cyber and influence operations

Anthropic also said it had identified Chinese and Yemeni groups attempting to use Claude to design conventional weapons. The company said it disrupted those efforts, as well as other activity involving data theft, malware and influence operations.
The company’s report described foreign groups using Claude to gather information about targets, identify vulnerabilities in online accounts, create malware intended to bypass security and steal data, cover their tracks, and distribute disinformation. “We’ve seen groups of actors use Claude to build networks of fake social media profiles and entire news sites,” the report said.

Anthropic said some of the campaigns showed hallmarks of Russian influence operations, including fake news sites and journalist profiles used to spread false claims about NATO activity in the Baltics. The company said U.S.-built AI tools streamlined those campaigns, increasing the volume of disinformation and helping create fake reporters that appeared more credible to local audiences.
Russian-linked groups, according to the report, “automated much of their operations from development, infrastructure acquisition, phishing, persistence through command and control, to data exfiltration.” One group used AI to forge documents attributed to the Central African Republic Gendarmerie and Ministry of Defense. Another, identified as GTG-84005, used AI to steal census and other public information, allowing agents to tailor messaging campaigns to audiences in Malaysia. Anthropic said the Turkish group also “laundered Russian and Chinese state media as independent reporting.”
A broader warning about AI-enabled influence
Anthropic’s report followed by two weeks a disclosure from rival frontier AI company OpenAI that Russian actors had used ChatGPT to create a fake “expert community” called the “International Burke Institute” to promote talking points to international audiences. Both companies said they had disrupted schemes they discovered, but the activity reflects a broader concern that such operations will continue as Russia uses AI to scale content aligned with Kremlin narratives, according to an August report from the Institute for the Study of War.
The primary source also reported that Russia is using young people’s interest in AI to expand the number of people involved in spreading Russian disinformation. Programs described as “AI media internships and training,” offered through university donations and related initiatives, are being used to reach young people around the world.
A recent article from New America similarly argued that AI gives authoritarian governments a powerful tool for distorting global understanding of geopolitical events. Daniel Kimmage, a former acting coordinator and principal deputy coordinator of the State Department’s Global Engagement Center, and Darjan Vujica wrote that “authoritarian regimes are early adopters in using AI to bend the arc of reality in their favor.”
The authors said the United States had previously established and operated organizations intended to counter such activity, but that those efforts had been undermined or halted during the second Trump administration. Less than a month after taking office, Attorney General Pam Bondi dissolved the FBI’s Foreign Influence Task Force, which investigated foreign disinformation and influence campaigns. Two months later, Secretary of State Marco Rubio shut down the Counter Foreign Information Manipulation and Interference hub, known as R/Fimi.
Kimmage and Vujica also pointed to the 2025 White House AI Action Plan’s call to remove references to misinformation. “The Trump administration has enthusiastically embraced AI-generated imagery in politically divisive posts,” they wrote.



