Sam Altman Lost the AI Jobs Narrative in May. This Role Is How OpenAI Plans to Get It Back.

OpenAI has posted a new role, a few weeks after Sam Altman walked back years of his own predictions about AI destroying jobs. At a conference in Sydney on May 26, Altman told the audience he was "delighted to be wrong" about how little damage AI had done to entry-level work. "I thought there would have been more impact on entry-level white-collar jobs being eliminated by now than has actually happened," he said. Independent research from Yale's Budget Lab and Brookings said the same this spring. Dario Amodei at Anthropic softened his own predictions in the same window. The OpenAI opening is for an Economist out of San Francisco. The team is called Economic Research. Read the responsibilities and it sounds like a job description at the Bureau of Labor Statistics.

This is not OpenAI's first move in this direction. In June, the company named Ronnie Chatterji, formerly Assistant Secretary of Commerce under President Biden and a professor at Duke's Fuqua School, as its first Chief Economist. A week later, on June 8, OpenAI launched the Economic Research Exchange, a formal program giving independent academic researchers structured access to OpenAI's proprietary usage data in exchange for research collaborations. The program's first named partners include Jason Furman, the former chair of the Council of Economic Advisers under President Obama, now at Harvard, and Michael Strain of the American Enterprise Institute and Georgetown. The Economist role posted last week fills out that team.

The role directs empirical research on employment, wages, skill requirements, and productivity effects of AI. It designs studies using proprietary company data alongside public government statistics. It publishes. It engages with philanthropies, agencies, and policymakers. It contributes, in the posting's own words, to "the development of new measurement approaches for AI's economic impact" and to "policy-relevant outputs, including academic papers, technical reports, and briefings."

The role serves two purposes. Internally, it feeds economic reasoning into OpenAI's product decisions. Externally, it produces academic papers, technical reports, and briefings that get handed to press, policymakers, and investors, and defines the frameworks other researchers use to measure AI's economic effect. The external side is where the heavier lifting happens.

A press release from OpenAI reads as marketing and gets discounted. A paper from a credentialed labor economist, even one on OpenAI's payroll, reads as research. It gets quoted. It gets cited. It becomes a data point that press coverage, Congressional testimony, and investor prospectuses all treat as neutral background.

The candidate profile is deliberately early-career. OpenAI wants a Ph.D. in economics with up to five years of post-Ph.D. experience, a labor, applied micro, or industrial organization background, and cross-functional research skills. This is the cohort of economists still choosing between industry, academia, and public-sector policy work, at exactly the moment they are making that choice. Anthropic, Meta, and Google DeepMind are recruiting from the same pool. So are the Fed's regional research divisions and the Council of Economic Advisers. OpenAI is bidding for a hire that will define its economic research output for the next five to ten years before the candidate ever tenures or reaches senior policy rank elsewhere.

That candidate cannot currently be recruited on academic pay. An assistant professor of economics at a top university earns roughly $150,000 to $200,000 with summer support. A junior economist at the Federal Reserve or the BLS starts around $130,000. Policy-shop analysts at Brookings or AEI fall in a similar range. OpenAI's ceiling is $385,000 in cash, with startup equity on top, decisively above what any statistical agency, university, or think tank could offer for the same early-career profile.

The gap is not accidental. It is the mechanism. OpenAI is not trying to compete for this hire in the AI-lab talent market. It is competing in the public-research talent market, and it is signaling that it can pay above what public institutions can afford. What OpenAI is buying is the credibility that goes with that credential, the same credibility that will show up in every paper, briefing, and policy conversation the researcher touches once they take the job.

OpenAI is also doing the opposite of what a pure narrative-capture strategy would suggest. The Research Exchange gives independent academic researchers structured access to OpenAI's proprietary data. Furman at Harvard and Strain at AEI, the two named collaborators, have institutional bases OpenAI does not control. That is not the profile of an operation that wants to own the narrative. It is the profile of an operation that wants to be present in every corner of it.

The 1990s pharmaceutical industry ran this same playbook. Rather than only fund external clinical research, the largest pharma companies built internal biostatistics teams that published in peer-reviewed journals under the company's byline. The papers were technically rigorous. The researchers were credentialed. For most of the 1990s the arrangement functioned as independent science. It took most of two decades for public trust to unwind, driven by the accumulated pattern of industry-employed researchers reaching industry-favorable conclusions at rates that could not be explained by chance. By the time the pattern was clear, the damage to public confidence in clinical numbers had been done.

That is the structural setup OpenAI is buying into. Not necessarily their intent. Their intent is likely a mix of genuine curiosity about what their products are doing and pragmatic need to answer the question in a form the outside world will accept. But the structural setup is what defines the long-run credibility of the resulting research, and that setup is essentially identical to the pharma playbook.

Which is why the most interesting question about the OpenAI role is not what OpenAI wants from it. It is what the person who takes the job insists on before they say yes. There are very few labor economists with the exact profile OpenAI is asking for. The ones who exist are being courted by every lab building the same kind of team. Whoever takes this role has real leverage right now. What that researcher demands in terms of publication independence, data access, peer review standards, and the right to publish findings the company would not have chosen to publish will do more to determine the credibility of AI economic research over the next decade than any framing OpenAI puts around the role.

If they demand a lot, this becomes a real research program that happens to be housed at OpenAI. If they demand little, it becomes a well-funded, credentialed narrative operation. The posting does not say which one it will be. Only the person who takes the job can decide that.

Time. "Sam Altman Says AI 'Jobs Apocalypse' Probably Won't Happen. What Changed?" Time, May 26, 2026. https://time.com/article/2026/05/26/sam-altman-ai-job-losses-openAI-/

Fortune. "Sam Altman and Dario Amodei Are Both Walking Back Their AI Jobs Apocalypse Prophecies as They Eye Blockbuster IPOs." Fortune, May 26, 2026. https://fortune.com/2026/05/26/sam-altman-dario-amodei-walking-back-ai-jobs-apocalypse-prophecies-ipo/

The Budget Lab at Yale. "Evaluating the Impact of AI on the Labor Market: November/December CPS Update." Yale Budget Lab, 2026. https://budgetlab.yale.edu/research/evaluating-impact-ai-labor-market-novemberdecember-cps-update

Brookings. "New Data Show No AI Jobs Apocalypse For Now." Brookings, 2026. https://www.brookings.edu/articles/new-data-show-no-ai-jobs-apocalypse-for-now/

OpenAI. "Economist." Ashby Jobs Board (OpenAI Careers), 2026. https://jobs.ashbyhq.com/openai/9ac67097-139b-4e3c-8c13-e9a6be26cf1b

OpenAI. "Dr. Ronnie Chatterji Named OpenAI's First Chief Economist." OpenAI Global Affairs, 2026. https://openai.com/global-affairs/openai-chief-economist-announcement/

OpenAI. "Introducing the OpenAI Economic Research Exchange." OpenAI, June 8, 2026. https://openai.com/index/economic-research-exchange/