OpenAI Proposes AI New Deal for 2028 Election Policy Agenda

OpenAI is making an early play to shape the 2028 presidential election conversation around artificial intelligence. In a policy essay published today, the company's chief global affairs officer Chris Lehane outlined five priorities he believes politicians must address, comparing AI's economic disruption to the conditions that led to President Franklin D. Roosevelt's New Deal in the 1930s. The timing follows a Democratic midterm sweep driven by voter frustration over cost of living issues, suggesting AI could become a central campaign theme as Americans worry about job displacement and rising electricity costs from data centers.

The Five Policy Priorities

Lehane's agenda spans worker protections, education reform, AI safety standards, data center development, and child safety measures. On the economic front, he proposes tax incentives for workers learning AI skills, easier capital access for entrepreneurs, and new approaches to employee ownership and profit-sharing. For education, he warns against keeping AI out of classrooms, arguing it could create inequality between wealthy students and children in rural communities. The policy framework also calls for enforceable AI safety standards, independent evaluations, public oversight, and the ability to use a kill switch to shut down dangerous AI systems.

Corporate Accountability Questions

Kate Brennan, who leads AI and corporate power policy at the AI Now Institute, told TechCrunch the proposals mark a notable shift from OpenAI's 2025 agenda, which focused on accelerating development and reducing regulatory barriers. However, she questions whether the New Deal comparison holds up. The original New Deal established worker protections and placed constraints on corporate power, while Lehane's vision offers few concrete mechanisms to ensure workers share in AI-generated wealth or that companies pay the costs of job displacement. Brennan suggested more binding measures would include revised corporate tax rates, incentives for hiring humans rather than automating jobs, and guardrails against market power concentration.

Data Center Development and China Competition

The essay acknowledges concerns about data centers while emphasizing construction jobs and local tax revenue they could generate. Lehane warns that politicians supporting a data center moratorium could jeopardize America's lead over China, comparing that stance to the 2002 Iraq war authorization vote. This framing comes after data center opposition emerged as a major bipartisan concern in several midterm races, with voters worried about rising electricity bills. Critics note this position aligns with OpenAI's infrastructure expansion needs rather than addressing community concerns about environmental and economic impacts.

Child Safety and AI in Education

Lehane makes child safety and education central to the agenda, calling for age-appropriate protections around AI companions for minors and warning against excluding AI from schools. The argument assumes AI improves learning outcomes, a claim that remains unproven according to critics who say more comprehensive research is needed before widespread classroom deployment. The policy also emphasizes maintaining human control over consequential AI-assisted decisions in education, work, healthcare, and government, though questions remain about what meaningful human control looks like as AI becomes increasingly required across these sectors.

Safety Standards and Past Opposition

While Lehane calls for enforceable safety standards and independent evaluations, Brennan characterized these stances as disingenuous given OpenAI's history. The company opposed California's SB 1047 AI safety bill in 2024, which would have mandated the catastrophic risk evaluations and shutdown capabilities Lehane now endorses. OpenAI also sought to limit state regulation in 2025, arguing states could undermine American innovation. The company has recently shifted some positions, backing several state safety laws and expressing openness to independent evaluators embedded in operations, though recent studies show OpenAI and other labs haven't published comprehensive plans for containing escaped models.

Release Date

The policy essay was officially released on November 4, 2026, positioning OpenAI's agenda two years ahead of the presidential election cycle.

What This Means

OpenAI's early policy positioning reflects the company's expectation that AI will become a defining issue in the 2028 election as affordability concerns intensify. Whether Lehane's proposals meaningfully redistribute power or entrench corporate control remains the central question. The framework presents AI as a path to broader prosperity for Americans willing to embrace the technology, but lacks binding mechanisms to ensure workers and communities share in the economic gains or protection from displacement costs. As AI becomes embedded across education, work, and government, the debate will likely center on who controls the technology's direction and who benefits from its deployment. OpenAI did not respond to requests for comment on implementation specifics by publication time.