Government and regulatory control over AI: national security, antitrust, and platform regulation
Key Questions
How has the Trump administration used national security authority on AI?
The administration pressured Anthropic over model capabilities, establishing a precedent for pre-release government review. The July 2026 roundup adds a voluntary pre-release framework and NSPM-11 for military AI.
What does the Great American AI Act propose?
It proposes preempting state AI laws to centralize federal control. This builds on existing tensions between national and state-level regulation.
What regulatory actions are competition authorities taking on AI?
The UK CMA imposed AI conduct requirements on Google, while the EU is pursuing DMA proceedings on AI access. Algorithmic collusion risks are also under active monitoring.
The Trump administration used national-security authority to pressure Anthropic over its model's capabilities, setting a precedent for pre-release government review. The July 2026 federal policy roundup adds a voluntary pre-release framework and NSPM-11 accelerating military AI. The Great American AI Act proposes state preemption, centralizing control. The UK government is pushing to require YouTube's algorithm to prioritize public service broadcasters (BBC), and YouTube is mobilizing creators to oppose the changes—a major test of algorithmic regulation and platform power. Competition authorities are actively enforcing: UK CMA imposed AI conduct requirements on Google, EU DMA proceedings on AI access, and algorithmic collusion risks are being monitored. A new article highlights the open-source vs. closed-model debate, with Box's CEO advocating for an 'alternative stack' to counter government control. Gary Marcus has raised concerns about LLM opacity and government overreach. A pragmatic opinion piece argues existing tools suffice for AI regulation.