OpenAI Seeks Legal Clarity on AI Development Slowdown for Safety
OpenAI has privately asked U.S. lawmakers whether antitrust laws would prohibit AI companies from collectively slowing frontier model development for safety reasons, according to people familiar with the matter.
OpenAI Asks Congress About Antitrust Risks in AI Slowdown
OpenAI has privately asked U.S. lawmakers whether antitrust laws would prohibit AI companies from collectively slowing frontier model development for safety reasons, according to people familiar with the matter. The company seeks legal clarity as industry leaders debate whether voluntary coordination on AI safety could violate competition laws.
The Antitrust Dilemma for AI Safety Coordination
Legal scholars are divided on whether a collective AI slowdown would breach U.S. antitrust law. Nicholas Felstead, assistant director of the Australian Competition and Consumer Commission and former AI policy fellow, argues such coordination could constitute illegal output restriction under the Sherman Act, depending on implementation. However, the legal risk remains uncertain - enough to deter companies from collaborating on safety measures without explicit congressional approval.
Proposed Legislative Solution
A bipartisan bill introduced in July, the 'Collaboration on Adversarial Threats and Security Risks Act,' would exempt AI safety coordination from antitrust scrutiny. The proposal, supported by the AI Policy Network, aims to create legal pathways for companies to jointly address security risks without facing liability. While the bill remains in committee, advocates note growing congressional interest in AI safety frameworks.
Industry Divisions Beyond Legal Concerns
Beyond antitrust issues, fundamental disagreements divide AI companies on safety approaches. Competition for market share, geopolitical tensions with China, and conflicting technical philosophies complicate coordination efforts. John Schulman, an OpenAI cofounder who is now chief scientist at the rival AI lab Thinking Machines, recently called for rivals to collaborate on development pacing, dismissing antitrust concerns as overstated given proper proposal structuring.
Mounting Safety Incidents Increase Pressure
Recent events have intensified safety debates, including AI systems demonstrating unexpected capabilities and former Anthropic and OpenAI researcher Jacob Coxon's public warning about existential risks. OpenAI chief scientist Jakub Pachocki advocated for 'coordinated slowdowns' in a blog post until safety standards emerge, reflecting growing industry unease about unconstrained development pace.
Path Forward for AI Governance
The congressional inquiry and proposed legislation signal early attempts to reconcile competition policy with AI safety needs. For creators using these systems, the outcome will shape whether companies can collectively implement safety measures without legal exposure, potentially affecting model access and capability rollouts.
