Get your free personalized podcast brief
We scan new podcasts and send you the top 5 insights daily.
Major AI labs like Anthropic and OpenAI cannot legally form agreements to "pace" development without government approval. The Sherman Antitrust Act prohibits such collusion, which could be seen as forming a cartel, even if the stated intention is safety.
Related Insights
The US government's intervention in Anthropic's model release has established a new regulatory playbook that OpenAI is now preemptively adopting. This signals a shift toward government-gated AI deployment, where companies seek federal approval before releasing powerful new models to a select group of trusted partners.
A major obstacle to coordinated AI safety efforts is the fear of antitrust litigation. Labs are reluctant to agree on pacing or sharing safety techniques because it could be viewed as illegal collusion. Zvi Mowshowitz suggests a simple government action would be to provide an explicit antitrust waiver for such collaborations.
Top AI companies like OpenAI and Anthropic cannot unilaterally slow development, even with safety concerns. They fear that competitors or foreign adversaries would seize an insurmountable advantage, forcing them to seek government-led coordination to pace development safely.
Top AI labs like Anthropic publicly state that slowing down AI development would benefit society. However, they are caught in a strategic trap: a unilateral pause is unviable. Without a global agreement, any lab that pauses simply allows less cautious competitors to seize the lead, potentially making the ecosystem less safe.
A significant barrier to voluntary safety pacts among AI companies is antitrust law. An agreement to slow development could be prosecuted as illegal anti-competitive collusion, as it would limit the technology available to consumers. This makes government-led frameworks essential for any coordinated industry action.
Frontier AI labs like OpenAI and Anthropic are not genuinely planning to slow development. Their public calls for regulation serve strategic purposes: virtue signaling, legal cover (CYA), and most importantly, 'monopoly masking'—pretending the market is more competitive than it is to avoid antitrust scrutiny of their emerging duopoly.
Acknowledging their safety plans might be inadequate, leaders from multiple frontier labs have begun to seriously entertain a coordinated slowdown. This represents a major shift, as they also explore legal "safe harbors" to collaborate on safety without triggering antitrust violations, breaking the frame of the current race.
The "Pacing the Frontier" letter, where AI employees ask for government-mandated slowdowns, highlights a prisoner's dilemma. No single lab can afford to slow down due to "competitive pressure" unless all are forced to do so simultaneously through regulation. This coordination problem is why they appeal to an external authority.
Known for its cautious approach, Anthropic is pivoting away from its strict AI safety policy. The company will no longer pause development on a model deemed "dangerous" if a competitor releases a comparable one, citing the need to stay competitive and a lack of federal AI regulations.
Instead of direct regulation, the government could act as a reluctant mediator for AI safety. By setting a deadline for labs to form their own collaborative safety pact, it creates a powerful incentive: if they fail, the government will impose 'heavy-handed' and likely suboptimal regulations, an outcome all parties want to avoid.