AI Lab CEOs Warn Their Own Companies Are Hiring Fewer Junior Workers

The leaders of Google DeepMind and Anthropic told a Davos audience on Tuesday that artificial intelligence is already reducing demand for entry-level employees at their own companies—a warning that carries weight precisely because it comes from the people building the technology.

“I can see it within Anthropic, where I can look forward to a time where on the more junior end and on the more intermediate end we actually need less and not more people,” Anthropic CEO Dario Amodei said during a joint interview with Google DeepMind CEO Demis Hassabis at the World Economic Forum. “My worry is as this exponential keeps compounding—and I don’t think it’s going to take that long, somewhere between a year and five years—it will overwhelm our ability to adapt.”

Hassabis echoed the concern, pointing specifically to internships. “I think we’re going to see this year the beginnings of maybe it impacting the junior level,” he said. Rather than pursuing traditional internships, Hassabis advised undergraduates to spend that time “getting proficient in learning these tools,” which he argued “could be better than traditional internships as you are leapfrogging yourself for the next five years.”

The remarks mark a notable shift. AI executives have spent years assuring the public that automation creates more jobs than it destroys. Now the CEOs of two leading labs are publicly acknowledging that their own hiring reflects a different reality—at least at the entry level.

The Data Behind the Warning

Amodei’s prediction that AI could eliminate half of entry-level white-collar jobs within five years, first made last year, remains unchanged. “Now I think maybe we’re starting to see just the little beginnings of it, in software and coding,” he said Tuesday.

Anthropic’s own research supports this trajectory. The company’s January 2026 Economic Index found that 49% of jobs can now use AI in at least a quarter of their tasks—up from 36% in early 2025. While the report emphasizes that AI augments rather than replaces most work, it notes that 45% of firms report reduced hiring needs at entry levels, even as only 9% report full role replacement.

The World Economic Forum’s Future of Jobs Report projects a net gain of 78 million jobs globally by 2030, with 170 million created and 92 million displaced. But that aggregate figure obscures stark differences by experience level. Entry-level job postings have fallen 29% since January 2024, according to Randstad analysis of 126 million postings worldwide.

For coding specifically—the field where AI assistance has advanced furthest—the numbers are starker. Programmer employment fell 27.5% since 2023. New graduate hiring at major tech companies collapsed by more than 50%. AI coding startups like Cursor have grown explosively, in part by helping fewer developers produce more output.

Orchestrators, Not Operators

Anthropic’s response to its own warning reveals what the new entry point might look like. Chief Product Officer Mike Krieger, the former Instagram co-founder, acknowledged that Anthropic has “tended less to hire fresh college grads” in recent months.

Instead, the company wants “people that are going to be defined more about the problems that they want to solve and how they can creatively solve them than a very specific ‘I know JavaScript and I’m going to do work in this exact environment,'” Krieger explained. Internally, Anthropic is shifting to hire experienced staff as “orchestrators of Claudes”—workers who direct AI systems rather than perform the tasks those systems now handle.

This represents a fundamental redefinition of entry-level work. The traditional on-ramp—learning by doing routine tasks under supervision—may be disappearing for roles where AI can perform those tasks faster and cheaper. The question is what replaces it.

The Broader Picture

The Davos warnings arrive amid broader anxiety about AI’s labor market impact. Morgan Stanley projects 200,000 European banking jobs will vanish by 2030, concentrated in back-office roles. Software stocks are tumbling as investors price in disruption fears following Anthropic’s Cowork launch earlier this month.

Hassabis struck a more optimistic note than Amodei, predicting “new, more meaningful jobs being created” and suggesting that any slowdown in internship hiring would be “compensated by the amazing tools out there for everyone.” He advised students to view AI proficiency as the new competitive advantage.

But both executives acknowledged uncertainty about the pace of change. Amodei said Anthropic is “thinking about how to deal with that within Anthropic in a sensible way.” What that looks like—for his company and the broader economy—remains unclear.

The World Economic Forum has launched initiatives including a Learning-to-Earning Sandbox that brings together universities, employers, and governments to redesign how education connects to employment. According to the WEF’s Davos briefing, 59 of every 100 workers will need reskilling by 2030, with 11 unlikely to receive it.

For now, the message from two of AI’s most influential builders is straightforward: the technology they’re creating is already changing who gets hired and why. Whether the economy adapts fast enough is a question neither could answer.