Government-mandated delays on public AI model releases, framed as a safety measure, do not slow internal development at major labs. This policy inadvertently creates a growing disparity between the powerful tools labs possess and what is available to the public, potentially making the AI ecosystem less safe and equitable.

While aimed at safety, proposed regulations could inadvertently create massive barriers to entry. The high cost and complexity of compliance would favor a few large, established AI labs, effectively giving them regulatory capture and stifling competition from smaller players and startups who can't make the cut.

By heavily policing large-scale compute clusters (the current path to AGI), regulations might inadvertently push researchers worldwide to secretly seek novel, resource-light paths to AGI that are harder to track and control, creating new risks.

A pause on training new, more capable AI models could paradoxically increase risk. It would halt progress at the few, relatively safety-conscious frontier labs, allowing less scrupulous competitors to catch up. Meanwhile, compute stockpiling would continue, making any subsequent capability leap even faster and more dangerous.

The US government's intervention with Anthropic's Fable 5 model signals a new era where AI labs will hold back their most capable systems from public release. This creates a consolidation of power, with only the labs and their chosen partners having access to true frontier capabilities.

The decision to restrict powerful but dangerous AI models like Claude Mythos to a select group of large corporations for safety reasons risks creating a massive centralization of power. This gives these entities an insurmountable technological advantage over smaller players and the public.

Tom Davidson cautions that an AI development pause, if implemented badly, could concentrate immense power in an unaccountable government body. This group could abuse its authority to approve or deny AI projects, ironically enabling a state-led power grab.

While nationalizing frontier AI seems like a control mechanism, it concentrates immense power within a potentially unstable political system. A more open, auditable, and decentralized AI ecosystem, despite introducing smaller risks, is argued to be more socially stable in the long run by diffusing control.

Slowing public releases of AI models for government review may not slow overall progress. This creates a scenario where labs advance internally for months, giving government agencies exclusive access while delaying public commercialization and the next cycle of investment.