OpenAI's Lobbying Proposal Under the Trump Administration: Accelerating Innovation and the Pursuit of AI Hegemony Under Relaxed Regulation

OpenAI's Lobbying Proposal Under the Trump Administration: Accelerating Innovation and the Pursuit of AI Hegemony Under Relaxed RegulationOn March 14th, following his return to the White House, President Trump swiftly overturned the first US executive order on artificial intelligence issued during the Biden administration. This move injected fresh uncertainty into the AI landscape and presented opportunities for companies like OpenAI to reshape the AI regulatory framework

OpenAI's Lobbying Proposal Under the Trump Administration: Accelerating Innovation and the Pursuit of AI Hegemony Under Relaxed Regulation

On March 14th, following his return to the White House, President Trump swiftly overturned the first US executive order on artificial intelligence issued during the Biden administration. This move injected fresh uncertainty into the AI landscape and presented opportunities for companies like OpenAI to reshape the AI regulatory framework. In response to this shift, OpenAI launched a lobbying effort, submitting a policy proposal to the US government on Thursday advocating for "accelerated innovation" and "relaxed regulation," while simultaneously highlighting the potential threat posed by China's AI advancements. This proposal targets the upcoming US "AI Action Plan," drafted by the White House Office of Science and Technology Policy (OSTP), expected to be submitted to President Trump before July.

Trump repealed the executive order on the "safe, secure, and trustworthy development and application of artificial intelligence," signed by Biden in October 2023, in January of this year. He subsequently issued a new order emphasizing the need to "maintain and strengthen America's global leadership in artificial intelligence" and demanding an AI action plan within 180 days. OpenAI's proposal is presented within this context. Notably, OpenAI recently secured a $40 billion investment from SoftBank, boosting its valuation to $260 billion. Under Trump's second term, OpenAI's position is delicate: it is deeply involved in Trump's "Stargate Project," a multi-billion dollar AI infrastructure investment, while simultaneously embroiled in a public relations battle with xAI, founded by Musk, a top Trump advisor. In February, OpenAI also announced plans to build data center campuses in 16 states.

OpenAI expressed dissatisfaction with current AI regulations in its proposal, arguing they stifle innovation. They called for "safeguarding the freedom to innovate in the national interest" and advocated for a "voluntary partnership" between the federal government and the private sector, rather than being constrained by "overly burdensome state-level laws." The proposal suggests a "fully voluntary and opt-in" collaboration mechanism between the federal government and leading AI developers and startups. Furthermore, OpenAI recommended that the US government implement "export control strategies" for domestically developed AI technologies to promote the global application of US AI systems.

To facilitate the rapid deployment and application of AI technologies, OpenAI proposed removing certain restrictive measures. Specifically, they suggested allowing federal agencies to "use real-world data for testing and experimentation" and a temporary exemption from the Federal Risk and Authorization Management Program (FedRAMP). They also proposed streamlining the process for AI companies to obtain federal security certifications, establishing a "fast-track approval process for AI tools based on clear standards," claiming this could expedite government adoption of new AI services by approximately 12 months. However, industry experts expressed concerns that rapid government adoption of AI could compromise safety and effectiveness, even raising questions about the necessity of the technology.

OpenAI further suggested government-private sector collaboration to develop AI technologies for national security, particularly emphasizing the need for "models trained on classified data and deeply optimized for national security tasks." These models should focus on "geospatial intelligence, classified nuclear missions, and other non-commercial areas." The launch of ChatGPTGov, tailored for the US government in January, exemplifies this strategy.

• On copyright issues, OpenAI recommended a copyright strategy that "promotes the freedom to learn," advocating for measures that "allow US AI models to learn from copyrighted materials." They argued that the "fair use doctrine" has fueled the development of US AI, emphasizing the crucial role of substantial investment and numerous startups in America's AI breakthroughs. However, since its public launch in late 2022, OpenAI's ChatGPT chatbot has been embroiled in multiple copyright infringement lawsuits due to its use of content directly extracted from news reports, including those from the Investigative Reporting Program, The New York Times, Chicago Tribune, New York Daily News, and numerous authors and visual artists.

OpenAI's proposal also prominently highlighted the perceived threat from Chinese AI technology, specifically targeting DeepSeek. OpenAI claimed DeepSeek is "state-subsidized" and "state-controlled," suggesting the US government consider banning models developed by such entities. However, there's currently no conclusive evidence of a direct link between DeepSeek and the Chinese government. DeepSeek was initially incubated by the quantitative hedge fund, Hudson Bay Capital Management. OpenAI also claimed that DeepSeek's R1 reasoning model poses security risks and could jeopardize user privacy because DeepSeek must comply with Chinese laws and provide user data upon request. It remains unclear whether OpenAI refers to DeepSeek's API, its open-source models, or both. In fact, DeepSeek's open-source models do not inherently contain mechanisms for the Chinese government to access user data, and companies like Microsoft, Perplexity, and Amazon host these models on their own infrastructure. OpenAI previously accused DeepSeek of violating its terms of service by "distilling" knowledge from OpenAI models; this proposal marks an escalation of OpenAI's crackdown on DeepSeek.

Worth noting is that in January, DeepSeek's application rapidly gained popularity in the US, surpassing ChatGPT in Apple's App Store rankings. DeepSeek's model development costs are a fraction of those incurred by competitors like OpenAI, Anthropic, and Google, raising concerns among American AI experts and investors. OpenAI stated in the proposal: "While the US currently maintains its lead in AI, DeepSeek demonstrates that our advantage is not large and is shrinking." OpenAI's proposal reflects not only its own interests but also the competitive landscape between the US and China in AI and the ongoing struggle over the future direction of regulation.