A paper released by researchers at the Foundation for American Innovation (FAI) found that artificial general intelligence (AGI) could either strengthen authoritarian regimes or destabilize democratic governance, depending on how it is adopted and controlled.

The March 2025 paper, AGI, Governments, and Free Societies, was authored by Samuel Hammond, Dan Lips, and Samuel DuBois. It warns that as governments incorporate AGI into their bureaucratic systems, the technology could enable levels of surveillance and centralized control never before possible. “Enhanced state capacity through AGI could enable unprecedented surveillance and control, potentially entrenching authoritarian practices,” the researchers wrote. “Conversely, rapid diffusion of AGI capabilities to non-state actors could undermine state legitimacy and governability.”

The authors examined how these risks play out at three levels of governance: the micro level of individual bureaucratic decisions, the meso level of organizational structure within government institutions, and the macro level of democratic systems as a whole. In each case, they argued that AGI could dramatically shift how power is exercised and accountability is maintained.

To mitigate these dangers, the researchers proposed a framework that includes robust technical safeguards, institutional designs that preserve meaningful human oversight, and regulatory systems capable of adapting to rapid technological change. “To preserve the narrow corridor of liberty,” the paper stated, “we propose a governance framework emphasizing robust technical safeguards, hybrid institutional designs that maintain meaningful human oversight, and adaptive regulatory mechanisms.”

The Foundation for American Innovation, previously known as the Lincoln Network, is a center-right think tank focused on the intersection of technology and governance. The organization promotes limited regulation, individual liberty, and policies that support technological innovation in a free society.