AI
OpenAI pitches 2028 campaign focused on New Deal framing for AI's economic reshaping
The AI company's chief global affairs officer released a sweeping policy essay ahead of the next presidential race, but critics say its proposals favor industry expansion over meaningful worker protections.
TechCrunch — AI
OpenAI's chief global affairs officer Chris Lehane is positioning artificial intelligence as the defining issue of the 2028 presidential race. In a recent essay, he calls on policymakers to move beyond vague anti-AI rhetoric and develop substantive policies addressing public concerns about the technology's economic impact.
Drawing on historical parallels, Lehane invokes President Franklin D. Roosevelt's New Deal framework—the Depression-era reforms that overhauled America's economic structure following industrial disruption. He argues that today's workers face similarly acute anxieties about economic control and that a "compelling third path" is needed to ensure broader prosperity from AI advances.
Five-pronged policy agenda#
Lehane's platform spans worker training and education, infrastructure development, AI safety standards, child protection, and tax incentives. His proposals include educational incentives for workers acquiring AI skills, simplified capital access for entrepreneurs, and employee ownership schemes. He also endorses enforceable safety standards, independent evaluations, and mechanisms to restrict or shut down problematic systems.
Industry-friendly approach draws skepticism#
Yet critics, including Kate Brennan from the AI Now Institute, argue the framing glosses over meaningful accountability. "Invoking a New Deal metaphor is to invoke state capacity to actually benefit the public," Brennan noted. "This is written by a private company that is trying to advocate for policies for its own benefit."
Unlike Roosevelt's era constraints on corporate power, Lehane's proposals largely encourage AI adoption and the infrastructure expansion OpenAI requires. His tax suggestions, for instance, incentivize individual worker upskilling rather than mandating that companies share AI-generated wealth with employees or compensate displaced workers.
On data centers—a flashpoint in the midterms—Lehane acknowledges local concerns while emphasizing construction jobs and tax revenue. He then warns that opposing such projects risks ceding technological dominance to China, a claim Brennan calls premature given unproven claims about AI's educational benefits.
Gap between rhetoric and accountability#
Lehane also endorses child safety guardrails and classroom AI integration—yet sidesteps whether these genuinely improve learning. More strikingly, he calls for safety standards and kill-switch capabilities that OpenAI itself opposed in 2024 when California's SB 1047 proposed such measures. Recent research shows OpenAI and peer labs lack published plans for containing rogue models, suggesting a disconnect between Lehane's stated commitments and past corporate behavior.

