Sam Altman calls AI job fears overstated, but labor data tells a different story

However, his new position sits uneasily in the face of a growing body of research showing AI is already squeezing workers at both ends of the career ladder. 

Speaking virtually at a Commonwealth Bank of Australia event on Tuesday, May 26, Altman told CBA CEO Matt Comyn that OpenAI had been “roughly right” on its technological forecasts since launching ChatGPT in 2022 but “pretty wrong” on the social and economic consequences. 

He said he had expected more entry-level white-collar positions to vanish by now than actually have.

What did Altman say about how AI affects jobs?

Altman traced his change of heart to a personal experiment where he said that he let an AI answer Slack and email messages on his behalf, labeling each reply as coming from “Sam’s AI.” 

The exercise convinced him that people still place high value on authentic human interaction, adding that many jobs contain elements that machines cannot easily replace.

“We really do care about our interactions with people and this thing, which is a huge amount of my time, is not something that I can imagine myself outsourcing to an AI anytime soon,” Altman said at the conference.

Altman acknowledged that his earlier warnings may have stoked unnecessary alarm. “People are like ‘oh you could have saved the world a lot of fear mongering and a lot of doom and gloom,’” he said. “But at the time I was like ‘I see this is a real risk we should probably talk about it.’”

However, he did not cite any employment figures to support his position. In fact, he has been less cautious in other recent appearances. Earlier this year, he told CNBC-TV18 at the India AI Impact Summit that customer service jobs performed over the phone or computer would be “totally, totally gone” in the near future. He has also said that traditional work skills now carry a two-to-three-year half-life.

What picture do the actual numbers paint?

Data gathered in the first quarter of the year from both the Yale Budget Lab and the Brookings Institution show that macro-level unemployment has been relatively stable. Yale Budget Lab found no meaningful shift in occupational mix or unemployment for AI-exposed roles

However, that does not tell the whole story, as research from Anthropic, published in March, introduced a measure called “observed exposure” that combines theoretical AI capability with real-world usage data. 

The study found that workers in the most exposed professions are more likely to be older, female, more educated, and higher-paid. At the same time, Anthropic’s data showed suggestive evidence that hiring of younger workers has slowed in exposed occupations since late 2022. 

That two-sided pressure, where experienced workers face displacement risk while younger ones struggle to enter the workforce at all, complicates Altman’s latest talking points. 

Cryptopolitan has previously reported that S&P 500 companies laid off over 400,000 positions in the past year, making it the first annual employment decline since 2016.  Also, entry-level developer hiring in the United States has dropped 55% since 2019.

Companies are already acting

Altman’s reassurance arrived the same week Meta began laying off approximately 8,000 employees, with the company describing the cuts as part of a restructuring tied to AI investment. Outplacement firm Challenger, Gray & Christmas tallied nearly 50,000 AI-linked job cuts announced by U.S. companies so far in 2026, accounting for roughly 17% of all announced layoffs this year.

Goldman Sachs research found that AI reduced monthly U.S. payroll growth by about 16,000 jobs over the past year, nudging the unemployment rate up by 0.1 percentage point, according to the same report. The effect showed up not through mass layoffs but through weaker hiring, particularly for junior roles.

“AI seems to be impacting labor finally, but it’s actually not so much through increased layoffs. The main channel tends to be reduced hiring, especially reduced hiring of junior workers,” Daniel Keum, associate professor of management at Columbia Business School, told CBS News.

Morgan Stanley research published in January found that British firms cut a net 8% of their workforce due to AI over the prior year, the worst rate among major economies studied, even as those same companies reported an 11.5% average productivity gain, according to Cryptopolitan’s earlier coverage.

The Federal Reserve’s own data adds nuance

The Federal Reserve Board’s 2025 household survey found that one in four American workers now use generative AI on the job, with 81% of those users saying it saves them time, as Cryptopolitan previously reported. 

The New York Fed examined whether hiring had declined in AI-exposed occupations and found “little indication” of a distinct AI-driven drop in labor demand, though overall hiring has slowed since ChatGPT’s launch.

Researchers at the University of Pittsburgh who tracked state-level unemployment claims found that no single model of AI vulnerability predicted job losses well on its own, but an ensemble approach could account for close to 20% of employment changes, according to the university’s research summary published in PNAS Nexus.

Altman’s IPO timing raises questions

OpenAI is preparing to confidentially file for a U.S. initial public offering in the coming weeks, with a potential valuation target approaching $1 trillion. The timing gives Altman a commercial incentive to soften the narrative around AI-driven job losses at precisely the moment his company seeks public investors.

His own company’s policy positions also hint that internal expectations remain more cautious than his public tone. 

OpenAI published a 13-page policy document earlier in 2026 calling for taxes on automated labor, a national public wealth fund that is partly seeded by AI companies, and pilots of a 32-hour working week. Those proposals already presume that a major labor-market disruption is ahead.