AI Mass Surveillance
Anthropic co-founder Dario Amodei has warned that AI could hand autocratic governments tools for total surveillance and control rawpixel.com/Magnific

Anthropic co-founder Dario Amodei has warned that increasingly powerful artificial intelligence could hand authoritarian governments the tools to build surveillance and propaganda systems so complete that citizens would have no way to resist or escape them. The warning appears in his essay 'The Adolescence of Technology', published on his personal website in January 2026.

In the essay, Amodei wrote that China has 'hands down the clearest path to the AI-enabled totalitarian nightmare I laid out above', pointing to Beijing's existing use of AI-based surveillance, including in the repression of Uyghurs. He argued that the risk he describes is not hypothetical, citing surveillance infrastructure and monitoring systems already in place today.

The Country of Geniuses Scenario

Amodei's essay sets out a scenario in which a country or company gains access to what he calls a 'country of geniuses in a datacentre', a term he uses to describe an AI system with the combined intellectual capability of tens of millions of highly skilled people. He argued that such capability, in the hands of an autocracy, could be used to read and interpret the world's electronic communications, identify dissent before it forms, and suppress it.

He also warned that AI-driven propaganda could be used to shape public opinion at a scale and precision no human-run system has managed before. This, he suggested, would make resistance far harder to organise than under previous generations of state control.

The essay was picked up beyond Anthropic's own circles. Foreign policy analyst Hal Brands cited it in a piece for Foreign Affairs, 'America's Superintelligence Dilemma', which examined how the United States might respond to the emergence of artificial superintelligence. Brands noted that Amodei has separately warned that superintelligent AI could cause a 'totalitarian nightmare' by enabling all-encompassing surveillance and automating repression.

The China Risk

Amodei's essay singles out the Chinese Communist Party as the actor he is most concerned about, citing its existing surveillance infrastructure and its use of AI-based monitoring in Xinjiang. He wrote that China is 'second only to the United States in AI capabilities' and has 'the greatest likelihood of surpassing the United States in those capabilities'.

He was careful to distinguish the Chinese government from the Chinese people, writing that he holds no animus toward China specifically. Ordinary citizens, he added, would likely suffer most from any AI-enabled repression, given they have no say in their government's actions.

Democracies Not Immune

Amodei's warning was not limited to autocracies. He wrote that democratic governments could also misuse AI surveillance tools against their own populations if existing safeguards erode, arguing that AI-enabled repression requires far fewer people to carry out than traditional forms of state control.

He called for governments to treat mass AI surveillance and AI-driven propaganda as 'bright red lines' that should not be crossed, regardless of who is doing the crossing. That standard, he argued, has to apply equally to allies and adversaries alike.

Amodei's warning matters because it comes from the leader of one of the world's most influential AI developers, not an outside critic. Anthropic's models are used by millions of people and by government agencies, giving his assessment of the technology's risks direct bearing on how it is built, deployed and regulated. That same fact cuts both ways, since Anthropic is itself a commercial supplier of AI to governments, a detail that has drawn scrutiny of Amodei's repeated public warnings about the technology his own company builds and sells.

If policymakers in Washington, London or Brussels treat his warning as credible regardless, it could shape upcoming decisions on AI export controls, surveillance legislation and international AI safety cooperation. Those decisions carry consequences that extend well beyond the tech industry itself.