OpenAI invokes Roosevelt's legacy: data-center battle precedes U.S. presidential election

OpenAI moved ahead of the midterm election results in the U.S. Congress, acting early to secure a seat in shaping the agenda for the 2028 presidential race. Chris Lihan, head of global affairs at the company, published an article urging policymakers to devise a third path that addresses voters' concerns about economic shifts, directly borrowing the "New Deal" model launched by President Franklin Roosevelt in the 1930s to restructure the economy after the Great Depression.

Lihan outlines five priorities: protecting workers, improving education, applying safety standards to systems, and expanding data-center infrastructure.This proposal marks a deliberate shift in the company's rhetoric, which has spent the past year focusing on accelerating development, loosening regulatory constraints, and facilitating physical expansion.The shift came after Democrats won the midterm elections amid widespread public discontent over living costs, inflation and electricity prices, concerns that also made opposition to data-center construction a shared electoral issue for both parties in several U.S. districts.

When we examined the proposal and its context, a deep paradox emerged, noted by Kate Brennan, AI policy and corporate power lead at the AI Now Institute. The original New Deal was built on using state power to curb large corporations and protect workers' rights, whereas OpenAI's document comes from a private firm advocating policies that primarily serve its commercial expansion. Lihan suggests tax incentives for employee training, easier financing for entrepreneurs and broader employee stock ownership, but avoids any legal obligations to share wealth or force companies to cover relocation costs, which Brennan sees as a pre-assumption that technology adoption will automatically benefit society without progressive taxes or concrete incentives to employ humans instead of automating their jobs.

The defensive tone is more pronounced in the data-center file; Lihan emphasizes construction jobs and local tax revenues, warning that any freeze on building would threaten U.S. superiority over China, even likening opposition to a decision to authorize the Iraq war in 2002. In the education file, he cautions against removing intelligent systems from classrooms to avoid widening the gap between city students and rural children, a reasoning that assumes, without solid evidence, that AI improves learning outcomes while ignoring expert calls to focus on supporting human teachers and directly funding schools.

The contrast extends to technical safety standards; the executive calls for binding standards, public oversight and the ability to trigger emergency shut-offs for models that lose control, while records show OpenAI opposed California's 2024 SB 1047 safety bill, which sought to impose the same catastrophic-risk assessments, sought to limit local legislation before later adjusting some positions and opening the door to independent auditors within its headquarters.

This political maneuver in Washington carries direct implications for financial and technical decision-makers in the Arab region, particularly in the Gulf, Egypt and the Levant.Turning data-centers and electricity consumption into an electoral battleground because of energy bills and living-cost pressures means regional alliances and investment funds that help finance U.S. computing complexes will face a more complicated political environment with increasing conditions and pressures to deliver tangible development benefits to local communities. At the same time, marketing AI tools in education as a way to narrow class gaps paves the way for a marketing push on ministries of education and regional institutions to purchase expensive software and subscription packages, even as facts confirm that operational priority should remain on training human personnel and developing realistic curricula before delegating the educational process to machines.

OpenAI's early attempt to influence the 2028 election reveals that the future of technological expansion is no longer just an algorithmic efficiency issue, but a struggle over resources and the distribution of public burdens. When AI firms ask societies to shoulder energy and grid pressures in exchange for promises of future prosperity, the real test shifts to public policy's ability to protect individuals' interests before monopolistic dominance takes hold behind slogans of national interest.