Anthropic says AI surveillance campaigns targeted dissidents and minorities
The company said it disrupted state-sponsored operations originating in China, Iran and West Africa between January and July, while alleging misuse of its models by Chinese developers.
Anthropic said it disrupted several state-sponsored surveillance operations that used its artificial intelligence models to target dissidents, minorities and diaspora communities. The campaigns, detected between January and July, originated in China, Iran and West Africa, according to a report cited by France 24 International.
The company said those targeted included Hong Kong pro-democracy figures, Tibetan and Falun Gong communities across Asia, and Iranian minority communities and opponents of Iran’s regime abroad. Anthropic said Iranian actors used AI to identify people through social media accounts.
Anthropic also said a contractor working for Malian national security authorities used Claude to design software that enabled intelligence gathering. The company warned that AI was increasingly being used in place of human engineering work for surveillance and other harmful activities.
The report alleged that Chinese developers at Moonshot and DeepSeek secretly used Claude to generate responses and distil data into their own models. Anthropic said Moonshot relayed almost 300,000 customer requests over 10 days through 5,380 allegedly fraudulent accounts, though it was unclear whether customers were told their requests were being rerouted to Anthropic.
Anthropic said it also blocked efforts involving weapons, scams and biological research. The biological cases involved pathogens, venoms and other toxins, but the company did not identify the researchers or assert that they intended harm.
The account is based on Anthropic’s report, with no independent corroboration provided in the source material.
