OpenAI proposes an agenda for the 2028 election

A political shift

Lehane urges policymakers to do more than issue anti-AI statements and oppose data center construction. They should offer people a way to share in the technology’s economic gains. In his view, Americans are once again worried about losing control of their economic future, as they were after industrialization and the Great Depression, when the New Deal introduced federal programs and reforms.

The essay appeared after a notable Democratic victory in the midterm elections: the party took control of Congress amid voter frustration over the cost of living. Kate Brennan, who leads work on AI policy and corporate power at the AI Now Institute, said Lehane’s current rhetoric differs from OpenAI’s agenda last year. At the time, the company argued for accelerating AI development, lowering regulatory barriers and expanding infrastructure.

Opposition to data center construction has become a prominent issue in several election campaigns, bringing together supporters of different parties. AI was not a leading issue in the midterms, but it could feature more prominently in political debates over the next two years.

Affordability is already emerging as a major issue in the 2028 election: Americans are facing inflation, tariffs and rising costs. AI could become a bigger part of that debate as concerns grow that data centers will raise electricity bills, automation will push people out of jobs, and a handful of wealthy companies will capture the economic gains.

How the benefits could be distributed

Lehane acknowledges these concerns but offers few specific measures to prevent AI’s benefits from becoming concentrated. His essay presents the technology primarily as a path to prosperity for Americans willing to use it.

On tax policy, Lehane proposes rewarding workers who develop AI skills, making it easier for entrepreneurs to access capital, and promoting employee stock ownership and profit-sharing. Brennan stresses that such incentives are not equivalent to binding rules that guarantee workers a share of the wealth created by AI or require companies to pay for the consequences of people losing their jobs.

Brennan also argues that a review of the corporate tax rate, tax incentives for hiring people rather than automating jobs, and measures to curb excessive market power would be more appropriate.

According to Brennan, Lehane’s proposals assume that widespread AI adoption will benefit society, though that has not yet been proven. She notes that every initiative in the essay is geared toward capturing the benefits of AI.

Data centers, schools and safety

On data centers, Lehane acknowledges the concerns but emphasizes construction jobs and tax revenue for local budgets. He warns that a construction moratorium could put US leadership at risk relative to China. Lehane compares politicians who support such a moratorium to those who voted in 2002 to authorize the Iraq War.

Child safety and education are also central to Lehane’s agenda. He supports age-appropriate protections and safeguards for minors using AI companions. Lehane also warns against excluding AI from schools: in his view, doing so could widen the gap between well-off students and children in rural areas.

That argument, however, assumes that AI improves learning. This has not yet been proven; critics say the question needs more careful study than simply increasing funding or giving teachers more individual attention.

Lehane calls for mandatory AI safety standards, independent evaluations and public oversight. He also proposes preserving the option to restrict systems or shut them down with an emergency switch. Recent studies, however, show that OpenAI and other AI labs have not published comprehensive plans for what to do if a model gets beyond human control.

Brennan called OpenAI’s position insincere and pointed to the company’s past actions. In 2024, OpenAI opposed California’s AI safety bill, SB 1047, which would have required assessments of catastrophic risks and provided for systems to be shut down. Last year, the company also sought to limit state-level regulation, arguing it could harm American innovation. OpenAI later changed some of its positions: it supported several state safety laws and said it was willing to let independent experts assess systems inside the company.

Lehane also argues that people should retain control over important decisions made with AI, rather than handing judgment entirely to machines. But the question remains how to ensure that control as AI becomes increasingly embedded in education, work, healthcare and government.

In his essay, Lehane describes AI as a technology that could restore Americans’ sense of agency and economic opportunity. At the same time, he ties the middle class’s success to participation in an economy increasingly dependent on companies like OpenAI. It is unclear whether the proposed measures would redistribute influence or strengthen the position of large companies.

TechCrunch asked OpenAI for more details about Lehane’s proposals and the company’s positions, but had not received a response by publication time.

Dataist’s take: The essay shows how OpenAI is trying to connect AI development to concerns about the cost of living, employment and political control. But a gap remains between the promise of broad public benefits and Lehane’s proposals: the measures he describes impose few binding requirements on companies, while AI’s benefits for education and the economy have yet to be proven.