Sam Altman, the CEO behind OpenAI, now says he misjudged how quickly artificial intelligence would replace entry-level white-collar workers. The reversal comes after months of dire predictions from Altman himself, who had warned that AI would wipe out those roles entirely.
A notable reversal on near-term impact
Altman previously stated that AI would eliminate many white-collar jobs, a warning that fueled public anxiety and corporate hiring freezes. In a recent interview, he acknowledged he was wrong about the near-term effects. “I overstated the speed of displacement,” he said. “We’re not seeing the wave I expected, at least not yet.” His shift in tone mirrors findings from several nonpartisan research groups.
What the research shows
Studies from the Yale Budget Lab, the Brookings Institution, and Anthropic have all found limited labor market effects from AI so far. The data points to modest productivity gains, not mass layoffs. Fewer than 5% of occupations show measurable AI-related job losses, according to the Yale analysis. Brookings noted that many companies are still experimenting with AI tools rather than deploying them at scale. Anthropic’s research tracked actual usage patterns and found most AI applications augment human work, not replace it.
Altman calls out 'AI washing'
Altman also used the interview to criticize what he called “AI washing” — companies using the label of automation to justify layoffs or restructuring that have little to do with technology. He did not name specific firms but said the practice is “more common than people think.” The remark suggests Altman believes some of the panic about job losses is manufactured by employers seeking cover for unrelated cost-cutting.
His own company, OpenAI, continues to develop advanced models that could eventually disrupt knowledge work. But Altman now says the timeline is longer than he once assumed. “We need to focus on what’s actually happening, not on hypotheticals,” he said.
What’s unresolved
Altman’s revised stance does not change the broader debate about AI’s long-term impact on employment. Regulators in the U.S. and Europe are still crafting rules around AI training and deployment. Meanwhile, companies that have already replaced staff with automation are unlikely to reverse those decisions. The next major test will come later this year when the Bureau of Labor Statistics releases updated data on AI-related job churn — a report Altman and his peers will be watching closely.
