Something unusual happened in the last week of May 2026. The two most prominent voices in AI, OpenAI CEO Sam Altman and Anthropic CEO Dario Amodei, both reversed course on the claims that helped put their companies on the map.

No more talk of “jobs apocalypse.” No more predictions of entire job categories being “totally, totally gone.” Instead: nuance, caveats, and a curious sense of relief.

Business leaders watching this space should take note, though for reasons different than you might expect.

What Altman Said

Speaking at a Commonwealth Bank of Australia conference in Sydney on May 26, Altman told CBA CEO Matt Comyn: “I’m delighted to be wrong about this. I thought there would have been more impact on entry-level white-collar jobs being eliminated by now than has actually happened.”

He went further, saying he had been “pretty wrong” on the social and economic implications of AI, even as he maintained the technological predictions he made at ChatGPT’s launch in late 2022 had been “roughly right.”

His explanation for the gap? He underestimated the value of the “human part” of work. After experimenting with delegating his own Slack and email responses to AI, he returned to doing them manually. People, he concluded, place real value on interacting with other people.

What Amodei Said

Amodei, who had previously suggested AI could eliminate 50 percent of white-collar jobs, also softened his position. He now says automation may actually expand the scope of work people do rather than eliminate it.

This is a notable shift for a CEO whose company has spent considerable effort warning about AI risk, including catastrophic economic displacement.

Why the Timing Matters

Both OpenAI and Anthropic are eyeing blockbuster IPOs in 2026, with each company carrying valuations approaching a trillion dollars. Fortune’s reporting notes that the shift in messaging coincides directly with preparations to attract institutional investors, who tend to be skittish about companies whose CEOs predict societal upheaval.

That does not make the updated positions wrong. But business leaders should factor in the incentive structure when evaluating what these CEOs say publicly.

What the Data Actually Shows

The picture is genuinely mixed. That is the honest answer.

On one side: the Yale Budget Lab published a study in May 2026 finding no significant changes in occupational mix or unemployment duration in jobs with high AI exposure since ChatGPT launched in late 2022. At the macro level, there has not been a wave of AI-driven unemployment.

On the other side: tech layoffs through May 2026 have passed 115,000 workers across more than 179 companies, already approaching the 124,000 logged across all of 2025, and running at a pace about 33 percent higher than the same period last year. Meta, Amazon, and Snap are among the companies explicitly citing AI as a reason for headcount reduction.

The macro unemployment data has not moved. But inside specific companies and job functions, the changes are real.

What This Means for Business

Here is the practical implication for business leaders: the extreme narratives (both the apocalypse version and the “nothing changes” version) are both wrong.

AI is not uniformly eliminating jobs at a macro scale. But it is changing what work gets done and how, and companies that are proactively adapting are already building structural advantages over those waiting to see how it plays out.

The businesses that come out ahead will not be the ones who panicked about AI taking all their jobs, and they will not be the ones who dismissed it as hype. They will be the ones who spent 2025 and 2026 building the data literacy, process discipline, and AI capability to actually use these tools well.

The fact that Altman and Amodei are walking back their apocalypse predictions does not mean AI’s impact on work is small. It means it is not playing out the way anyone expected. The thoughtful, practical approach always beats the dramatic one.

A few questions worth asking inside your own organisation right now:

Which roles in your team currently spend more than 30 percent of their time on tasks that AI handles well today? Not to eliminate those roles, but to free those people up for higher-value work.

Do your managers know how to evaluate AI-generated output? The risk is not that AI replaces workers. The risk is that managers lack the judgment to know when AI output is wrong.

Are your processes documented well enough for an agent to run them? AI agents amplify well-structured workflows and expose poorly-designed ones. If your processes live in people’s heads, you cannot scale with AI, full stop.

The jobs apocalypse is not here. But the competitive pressure to build AI capability absolutely is.

---

Enterprise DNA helps businesses build the data literacy and AI readiness to compete in this environment. Talk to us about where your team stands.