Supreme Court Decisions Limit Federal Options for Open-Weight AI Oversight
The article argues that current federal regulation of AI relies on threats against frontier companies, but open-weight models will be harder to control. Supreme Court rulings have narrowed the government's regulatory authority. A new agency or expanded role may be needed but faces legal obstacles.
The current federal oversight mechanism relies on presidential pressure—export controls and supply-chain risk designations—against a small circle of closed-weight frontier firms like OpenAI and Anthropic. These companies seek favor with the administration, making threats effective. The article notes that the legality of such actions is often secondary to the president's willingness to use retaliatory leverage unrelated to model safety.
Open-weight models, freely copyable and modifiable, will soon match frontier capabilities. Once released, numerous small entities, including U.S. firms like Fireworks AI offering China's Kimi K3, can replicate them without federal contracts. Supreme Court rulings since the 1990s have narrowed federal regulatory authority, creating a legal gauntlet for any new agency or expanded existing role to address existential AI risk.
The shift to open-weight models could fundamentally alter the balance of power in AI governance. If federal regulators lack legal tools, dangerous capabilities may spread beyond any single nation's control, affecting global security and public trust. Conversely, an overreaching new agency could chill innovation and civil liberties. The Supreme Court's constraints may force a choice between inadequate oversight and legally dubious executive action, leaving citizens and companies in a precarious position.