OpenAI Details How It Would Like AI to Be Regulated

OpenAI, the Sam Altman-led company that ushered in the AI era with the late 2022 release of ChatGPT, has laid out its idealized vision for how the United States government, and the incoming Trump administration, can grow and regulate the country’s burgeoning artificial intelligence industry. In a document titled “AI In America: OpenAI’s Economic Blueprint,” the company calls for infrastructure investments, collaboration between private AI companies and government agencies, and a light touch of federal regulation, rather than letting each state decide their own rules. In an introduction, OpenAI vice president of global affairs Chris Lehane compared the current state of the AI industry with the early days of the automotive industry. He pointed out that the United Kingdom’s auto industry was stunted by overregulation, while the United States became the car capital of the world by “merging private-sector vision and innovation with public-sector enlightenment.” Lehane wrote that the U.S. has another chance to be a leader in a potentially massive market, but only if the government and industry can work together. “In the same way the federal government helped clear the way for the nascent automobile industry to grow, including by preempting a state-by-state tangle of roads and rules,” wrote OpenAI, “it should clear the way for the AI industry’s development of frontier models.” The context behind this document is that AI companies are increasingly worried that, without any federal regulation, individual states will develop their own rules. In 2024, California governor Gavin Newsom vetoed a bill that would’ve instituted safety requirements for AI models, but did sign a bill mandating the disclosure of training data for AI models. One solution proposed by OpenAI is to develop pathways for companies that develop large language models to have their models evaluated by a government agency. “In return,” the proposal says, “these companies would receive preemption from state-by-state regulations on the types of risks that the same national security agencies would handle.” (Notably, OpenAI does not propose any sort of government evaluation that AI models would need to undergo in order to be publicly deployed.) But state governments still have a big role to play in OpenAI’s grand vision. OpenAI proposed that state and local governments create “AI economic zones” in order to speed up the permitting process for building both data centers and power-generating infrastructure like wind farms, solar arrays, and nuclear reactors. The company estimates that there’s “$175 billion in global funds waiting to be invested in AI infrastructure,” and if the United States doesn’t take those funds and build infrastructure here, the Chinese Communist Party will happily step in. OpenAI is embarking on a nationwide “Innovating for America” initiative to sell its vision at the federal, state, and local level, starting with a Washington, D.C. event on January 30, in which CEO Altman will preview new OpenAI technology and discuss its ability to drive growth. It still remains to be seen how Trump’s relationship with OpenAI rival Elon Musk will impact the administration’s reaction to these proposals, but one thing is certain: Whatever the Trump administration decides to do with AI regulation will have major repercussions for businesses that build and use AI models. BY BEN SHERRY @BENLUCASSHERRY