Chris Lehane, OpenAI’s chief global affairs officer, believes that the 2028 presidential election could significantly impact the future of artificial intelligence (AI). He stresses the need for policymakers to create strategies that genuinely address public concerns regarding the technology rather than merely offering shallow responses to growing anti-AI sentiment.
In a recent essay, Lehane likens the transformative potential of AI to the New Deal initiated by President Franklin D. Roosevelt in the 1930s, which aimed to reform a decimated economy following the Great Depression. He notes that, similar to that era, many Americans today are feeling uncertain about their economic stability and calls on political leaders to adopt a “compelling third path” that allows for broader access to the advantages of AI. His essay outlines five key policy areas that include worker protections, educational initiatives, AI safety, and the development of data centers.
This essay arrives on the heels of a significant Democratic win in the midterms, where the party regained control of Congress amidst public dissatisfaction over rising living costs. Kate Brennan from the AI Now Institute remarked that Lehane’s current stance marks a stark departure from OpenAI’s previous focus on fostering AI development and diminishing regulatory obstacles.
Many of Lehane’s suggestions appear to advocate for enhanced AI integration while also supporting the infrastructure needed for companies like OpenAI to expand. The original New Deal focused on worker rights and imposed limits on corporate power. However, it remains ambiguous how Lehane’s vision would hold AI firms accountable for the potential economic disruptions stemming from their innovations.
“Using a New Deal analogy suggests a governmental role in promoting public good,” Brennan stated. “This perspective is coming from a private entity aiming to champion policies that could benefit its own interests, which is troubling.”
Although AI was not a key issue in the midterms, the opposition to data centers became a significant bipartisan topic. Lehane’s assessment that AI will gain prominence in upcoming discussions appears valid as concerns surrounding job security, rising electricity costs, and economic inequity grow among voters.
While acknowledging these apprehensions, Lehane does not present many actionable plans to ensure the equitable distribution of AI’s economic benefits. Instead, he emphasizes AI as a means for enhanced prosperity for those who embrace it.
Lehane proposes various changes to the tax system, including providing incentives for individuals to acquire AI skills, facilitating easier access to capital for entrepreneurs, and introducing new profit-sharing mechanisms. However, these proposals do not equate to compulsory measures ensuring that workers benefit from AI-driven profits or that companies address the consequences of worker displacement.
“I found much of this policy framework to be aimed at a future dependent on widespread AI adoption benefiting society, which remains unverified,” Brennan noted. “All of it assumes AI’s positive potential.”
Brennan suggests that more sensible approaches might involve restructuring corporate tax rates or providing incentives for hiring human workers instead of resorting to automation, alongside measures to curb excessive corporate power.
Lehane’s focus on expanding economic prospects through AI rather than addressing its possible negative ramifications also extends to data center discussions. He mentions the community benefits these centers could bring while cautioning that those advocating for a moratorium on them risk compromising America’s competitive edge against China, likening this to the controversial 2002 Iraq war vote.
He further emphasizes the significance of child safety and education in his policy vision, advocating for appropriate protections around AI companionship for minors. Lehane warns that excluding AI from educational settings might exacerbate disparities between affluent students and those from rural backgrounds. However, this assertion presupposes the effectiveness of AI in learning, a claim yet to be substantiated and one that critics argue needs rigorous scrutiny rather than simple funding increases or more individualized teacher attention.
Lehane calls for enforceable safety standards for AI, independent assessments, public oversight, and mechanisms to deactivate AI systems. However, recent studies have indicated that organizations like OpenAI have not established clear protocols for addressing potential loss of control over AI models.
Brennan criticized these views as “disingenuous,” highlighting OpenAI’s historical resistance to legislation aimed at enforcing such safety measures. For instance, in 2024, OpenAI did not support a Californian bill that would have required evaluations for catastrophic risks associated with AI. In the past, the company also contested state regulations, claiming they could hinder American innovation. Recently, however, OpenAI has shown some willingness to align with various state safety laws.
Lehane believes that individuals should retain authority over critical decisions influenced by AI rather than relinquishing judgment to machines. This raises questions about what genuine human control entails in a world where AI use is becoming increasingly common in sectors like education, healthcare, and governance.
Throughout his essay, Lehane portrays AI as a tool to restore economic opportunity and agency for Americans. However, he connects the success of the average American to their engagement in an economy increasingly dominated by companies like OpenAI, leaving uncertain whether his proposed policies will effectively redistribute power or subtly reinforce existing inequalities.
Inquiries were made to OpenAI for further details on Lehane’s proposals and the organization’s positions, but no response was received before publication.
