Special Edition: How to Win as an AI Populist

AI for DC | By Chris Lehane

---

Hi

Welcome (back) to The Prompt. AI isn’t well understood, but we learn a lot in our work that can help. In this newsletter, we share some of these learnings with you. If you find them helpful, make sure you’re signed up for the next issue.

This special edition of The Prompt is a personal essay from OpenAI Chief Global Affairs Officer Chris Lehane on advice he’s giving candidates in the 2026 US midterm elections.

---

A Fair Chance Agenda for 2026*

A growing number of candidates have reached out to me asking the same question: what should a 2026 AI agenda look like? Based on the 2025 outcomes, their own polling, and what they hear from voters and press, they know AI will be on the ballot with them, and they need an AI agenda that solves for three realities:

1. The US must keep winning the AI race with the People’s Republic of China. No party or politician can be in the position of having lost AI to the PRC.

2. America must continue to benefit from AI’s upsides. We’re already seeing better healthcare outcomes, faster science, and real gains for small businesses.

3. Americans themselves need to win when it comes to AI and their jobs, their kids’ futures, their kids’ safety, and their cost of living.

These candidates, both incumbents and challengers, are calling not just because I’ve run and advised campaigns for a long time, but because they’re frustrated by an oversimplified debate disconnected from reality and voters’ lives. It leaves them little to work with during long campaigns in which they need to address both the aspiration and the anxiety around AI.

On one side, accelerationists argue not to worry, the market will sort it out: some disruption now but don’t sweat it, soon we’ll have a world full of cheap goods, powerful robots, and work being optional, like playing sports or a video game. Not exactly the lived experience for most Americans.

On the other side, doomer elites argue that AI will wipe out 20% of jobs (and yet somehow help shrink the national debt), so it should be tightly controlled by a small group of establishment decision-makers. Guess who? On something this consequential, everyone should have a voice, not just an autocratic few.

Voters don’t believe the accelerationists and don’t want to believe the doomers. Candidates don’t want to be stuck choosing between selling a fantasy or scaring people. All are looking for a third option—one where voters have a voice and leaders are listening and leading. Voters want to know how AI will work for them, not just around and past them. Nothing good comes from Americans feeling unheard.

Good AI policy is good politics. Majorities agree we must beat the PRC on AI and that investing in AI is investing in America. At the same time, they worry about their jobs, their kids, and rising costs. They want leaders who can explain how families can participate and prosper with AI, not just survive it.

Campaigns win by playing offense with ideas that look forward, speak to the moment, and earn trust. Durable majorities are built by offering real solutions grounded in real values. AI is moving too fast for small thinking. Voters know it, and candidates who don’t stake out their own AI agendas will be trapped between hype and fear.

So here’s what I’ve shared, in my personal capacity, with candidates who have asked.

The stump speech

For millions of Americans, wages have lagged, costs keep rising, and too many feel one bad break away from falling behind. That was true before AI. Now AI is accelerating everything and forcing a hard question: does our economy work for those who actually do the work?

AI is moving fast but its benefits aren’t. A small group is pulling ahead, while everyone else risks being left behind. If we do nothing, the economy might grow but working people will keep losing ground. That’s not progress. And it’s not inevitable.

Today’s debate is dominated by two loud extremes. One wants to charge ahead without regard for real-world harm, asking you to trust that the invisible hand will fix things later. Communities that lived through the China shock aren’t buying that again.

The other wants to slow or stop AI and limit access to elites, experts, and regulators. That’s a lot like Europe’s approach, and is one that leaves countries further behind.

Both paths fail working people. Both entrench power. Both reduce your agency to decide for yourself and act on your own.

There is a third path that aligns agency with growth. At our best, since our founding 250 years ago, America has been a nation of builders, creators and learners. With AI, we can build and create much bigger and learn much more by lowering barriers to entry and equipping people to turn ideas into livelihoods.

We should advance AI in the open while scaling people’s ability to understand, use, and benefit from it. With access and education, anyone can use these tools to create opportunity. Done right, AI can renew a deeply American idea: balancing between capital and labor so everyone has a fair chance.

The US leads in AI and must continue to outcompete China, but we can only win globally if Americans are winning at home. Here’s how we make that real.

The big ideas

1. Make AI a basic right. Before 2030, every American should have guaranteed access to a baseline level of AI. AI literacy should be as fundamental as reading, writing, and math. Lifelong learning and certification must be widely available so people can adapt as the technology evolves. If AI is the printing press of our era, everyone needs to be able to read. China is aiming for nationwide AI education by 2030. We need to beat that, and close the participation gap before we repeat the China shock, only worse.

2. Set a goal for a “participation economy.” Just as we track inflation and unemployment, we should set a national goal linking AI use, and the economic growth that will be generated by it, to a good middle-class life. Make $147,000 (about the income needed for a secure middle-class life) a real target and not a Reddit argument, and make policy that gives families a fair chance to participate in an AI economy. Power users of AI tools are seeing real productivity gains, meaning they are producing more economics for their employers; they – not just their companies or executives – should share in the value they create.

We should push AI companies to share real-time data on productivity gains so we know where to support and pay workers more. In an AI economy, outdated licensing rules make no sense except as a barrier to the opportunities AI can unlock for working Americans. With modest reforms, a union electrician could do engineering work on a home build, or a home health worker could deliver primary care. We can have fewer billable hours for lawyers and better-paid teachers in classrooms.

3. Better align corporations with the public. To modernize corporate responsibility for the AI era so that public interests are represented as companies grow, build on existing public-benefit structures to ensure that frontier-model companies help fund training, transition, and shared prosperity. Ideas like newborn savings accounts or dedicating profits to retraining are good starts, but we should go further. Imagine a new corporate structure where the public holds roughly one-third ownership, alongside investors and employees. There’s a lot of talk about a billionaire’s tax, but if we’re serious about aligning the public interest with AI’s returns, let’s really get serious so that when AI companies win, Americans win, too. To those who call this pie-in-the-sky thinking, consider OpenAI’s funding one of the largest nonprofits in history by allocating about 30% of the business to a foundation that also governs the company.

4. AI knows a lot, but parents know best. Kids’ safety must come first. Parents need real tools and real control. That means doing what we didn’t do in the social media era: requiring that companies verify a user’s age. It means time limits, no companion bots for kids, and building kids safety into a single national overall AI safety standard that avoids loopholes. We also need AI tools that help kids learn, whether they grow up in rural Maine or Silicon Valley.

5. Treat infrastructure as destiny. We can’t win the AI race, or give Americans a fair chance in the AI economy, without massive investment in energy and data infrastructure. In 2024, China added about 440 GW of energy capacity; the US added about 50. AI infrastructure should lower electricity costs, create union jobs, strengthen grids during emergencies, and accelerate clean and reliable energy from geothermal to nuclear. Not even Sen. Bernie Sanders could oppose data centers that cut costs, modernize our sources of energy, create union jobs, and help beat the CCP.

6. Transform healthcare and science. AI can make healthcare better, cheaper, and more accessible, rendering old public-versus-private fights obsolete by changing the cost structure entirely. Families in healthcare deserts already rely on AI for better outcomes and to navigate insurance, sending over 500,000 healthcare-related messages every week from these underserved areas. Laws should reflect that reality so every family can use AI to improve their health now.

The Fair Chance Agenda is about advancing the tech, increasing people’s agency, and sharing the gains. That’s how America has always won, and how Americans can win again.

* I’m big on fairness when it comes to putting people before the powerful. Fair deals, fair chances, fair shakes, fair shares. Today’s revolutionary technologies can create fairness where it doesn’t currently exist, and too often, we see that potential get misdirected. Let’s not let that happen with AI.

---

[About] Launching OpenAI Academy for News Orgs

The Prompt’s bench of authors and contributors is loaded up with former journalists, and we routinely think about how we would have incorporated AI and ChatGPT into our old jobs. So we’re extra psyched about the launch of OpenAI Academy for News Organizations, in partnership with the American Journalism Project and The Lenfest Institute for Journalism. It’s a hub for journalists, editors, and publishers using AI.

At the outset, this Academy will provide on-demand training (including “AI Essentials for Journalists”), practical use cases, open-source projects, and guidance on responsible uses.

[Disclosure]

Graphics created by Base Three using ChatGPT.

---