OpenAI Urges Trump Administration to Adopt AI Action Plan by July 2025, Proposing Federal Preemption and Fair Use Codification

OpenAI has submitted a comprehensive policy proposal to the Trump administration, advocating for a strategic approach to artificial intelligence (AI) regulation as part of the forthcoming 'AI Action Plan,' due by July 2025. The proposal emphasizes the need for swift AI advancement and minimal regulatory constraints, highlighting concerns about competition from China's AI sector, particularly from the startup DeepSeek. The company argues for federal preemption of state AI laws, suggesting that AI companies should be shielded from state regulations if they voluntarily share their models with the federal government. OpenAI also calls for an 'export control strategy' to promote the global adoption of American AI systems while restricting technology transfers to China. In terms of copyright, OpenAI urges the U.S. government to codify 'fair use' for AI training, asserting that the current doctrine has been crucial for the growth of AI startups and research in the U.S. This stance comes amidst ongoing legal battles over copyright infringement related to AI training data, including lawsuits from major news outlets and creators. OpenAI's proposal also addresses national security, recommending that the government partner with the private sector to develop AI for specialized tasks, such as geospatial intelligence and nuclear operations. The company suggests that federal agencies should be allowed to experiment with AI using real data and proposes a faster approval process for AI tools in government use, which could allow access to new AI services 12 months earlier. OpenAI, currently valued at $260 billion with a pending $40 billion investment from SoftBank, is in a competitive landscape with Elon Musk's rival AI startup, and has considered building data center campuses in 16 states. The company has also released ChatGPT Gov, a product specifically designed for U.S. government use.