risk · R09

AI-Enabled Mass Surveillance and Authoritarian Governance

The threat of AI-enabled surveillance is rapidly expanding and diversifying. While state-led initiatives and the export of authoritarian models continue, major technology companies are building an increasingly ambitious privatized surveillance infrastructure. Recent examples include plans to turn gig-economy drivers into mobile surveillance fleets and patents for consumer smart glasses with real-time facial recognition. This convergence of state and corporate power, coupled with the export of censorship norms via commercial AI models, creates an unprecedented and accelerating risk to privacy and civil liberties globally.

Updated 2026-10-25

Observatory assessment: 95 / 100. Critical

How assessments are made

Deep dive

AI has fundamentally shifted the balance of power between states, corporations, and citizens. While the state-centric model, exemplified by China's integrated surveillance infrastructure, continues to be exported globally, a parallel threat is emerging from the commercial sector. Western tech giants like Meta, Amazon (Ring), and OpenAI are developing and normalizing consumer surveillance devices such as smart glasses and home assistants with facial recognition. This creates a decentralized, peer-to-peer surveillance network that complements state capabilities and erodes privacy in previously private spaces like homes and workplaces. The '100% identification' capability of modern facial recognition combined with real-time analysis creates chilling effects on free speech, assembly, and dissent, with the threat now stemming from both government and corporate actors. The lack of international norms and domestic regulation on consumer surveillance technology accelerates this trend, enabling a pervasive surveillance society even in democratic states.

Trajectory and projections

Near term

By 2026, AI-powered surveillance is standard in 100+ countries, driven by both state adoption of 'turnkey authoritarianism' packages and the rapid expansion of corporate surveillance networks (e.g., ALPRs, smart glasses) in democracies.

Mid term

By 2030, real-time emotion recognition in public spaces enables 'pre-crime' policing in several nations. Democratic pushback results in surveillance moratoriums in some EU cities.

Long term

By 2035, the global landscape is split: surveillance-free zones in some democracies vs. total information awareness states. The UN debates a 'Right to Anonymity' convention.

Mitigation strategies

• International moratorium on AI facial recognition in public spaces
• Export controls on surveillance AI technology to authoritarian regimes
• Mandatory transparency reports for government AI surveillance programs
• Support for privacy-preserving technologies and encrypted communications
• Strong consumer data privacy laws regulating data collection by AI-enabled devices
• International legal challenges against state surveillance (e.g., via human rights courts)
• Local government moratoriums on specific surveillance technologies (e.g., ALPRs, facial recognition)
• Judicial warrant requirements for police searches of stored ALPR and automated location databases
• Mandatory independent auditing and strict access controls to prevent officer misuse of surveillance systems

Top evidence