OpenAI releases vision for a new AI policy agenda, positioning early for the 2028 US election issues

In a newly released policy article, Lehan compares the current social and economic changes brought about by artificial intelligence to the period of industrialization the United States went through in the 1930s. He argues that Franklin D. Roosevelt reshaped America's social contract through the "New Deal" at that time, and that the AI era likewise needs a new policy framework to help the public adapt to change.

Lehan points out that many Americans today are worried about losing control of their economic future, and are concerned about the employment impact of AI, imbalances in wealth distribution, and changes in social structure. He advocates that political leaders should seek a "third way"—neither allowing AI to develop without any limits, nor letting technology be monopolized and controlled by a few, but rather enabling more people to share in the benefits AI brings.

To this end, OpenAI has proposed five key policy directions, covering areas such as worker rights protection, education reform, AI safety, data center construction, and child protection. These include encouraging workers to learn AI skills, helping entrepreneurs obtain more funding support, exploring profit-sharing mechanisms, and promoting broader integration of AI into the education system.

In the field of education, Lehan warns that if schools completely reject AI tools, it could further widen gaps in educational resources. He believes AI should be incorporated into the future education system rather than simply excluded from the classroom.

At the same time, OpenAI also calls for establishing more comprehensive AI safety standards, including independent evaluation mechanisms, public oversight systems, and emergency measures to shut down dangerous AI systems when necessary.

The timing of this policy vision's release is politically significant. Not long ago, the U.S. midterm elections had just concluded, with Democrats successfully regaining control of Congress, and voters' dissatisfaction with rising living costs, inflation, and economic pressure was seen as an important influencing factor. Against this backdrop, the risk of job displacement brought by AI and the issue of economic benefit distribution are becoming increasingly important public issues.

Kate Brennan, policy lead on AI and corporate power at the AI Now Institute, says Lehan's remarks this time differ markedly from OpenAI's policy priorities over the past year. Previously, OpenAI focused more on accelerating AI technology development, reducing regulatory barriers, and promoting infrastructure construction, whereas now it has begun to place greater emphasis on showing the public the potential social value of AI.

However, critics argue that this "AI New Deal" vision is still fundamentally beneficial to OpenAI and the entire AI industry. Brennan points out that Roosevelt's New Deal not only promoted economic development but also constrained corporate power and strengthened worker protections. By contrast, OpenAI's proposal involves relatively little on how AI companies should bear responsibility for potential future economic shocks.

She believes that to truly safeguard the public interest, policymakers also need to consider more binding measures, such as adjusting corporate tax systems, encouraging companies to retain human jobs, limiting excessive market concentration, and ensuring that wealth created by AI can be returned to society more broadly.

Although AI was not the most central topic in the 2026 midterm elections, disputes over data center construction have already begun to emerge in multiple constituencies. Some residents worry that large AI data centers will drive up electricity consumption, electricity prices, and resource pressure, while the tech industry emphasizes that these projects can bring jobs and economic growth.

Industry analysts believe that as AI capabilities continue to improve and penetrate all areas of the economy and society, discussions around employment, energy, education, and wealth distribution will become increasingly frequent. By the 2028 presidential election, AI is likely to no longer be merely a topic for the tech industry, but will become an important political issue that concerns the vital interests of ordinary voters.

From the current perspective, OpenAI clearly does not want to play only the role of a technology provider. By releasing policy positions in advance, influencing the direction of political discussion, and promoting a regulatory framework conducive to industry development, this AI company has already begun actively participating in shaping the future model of AI governance in the United States. And the debate over how AI should be integrated into society, who should benefit from it, and who should bear the risks is also very likely to become one of the focal points on the American political stage in the coming years.