The $900 billion reason Anthropic CEO Dario Amodei keeps talking about AI taking away millions of jobs
Dario Amodei has spent the better part of 2026 warning anyone who will listen that AI is about to wipe out half of all entry-level white-collar jobs. He said it at Davos in January. He said it in a 20,000-word essay published the same month. He said versions of it at the India AI Impact Summit in February, and elaborated on it in an interview shortly after. The message barely changes: coding jobs go first, then finance, law, and consulting follow, and the displacement could be “unusually painful.”Unemployment, in his telling, could hit 10–20% within five years—which, applied to the global knowledge workforce, would amount to hundreds of millions of people out of work.It’s a striking thing to say, especially for the CEO of the company building the very AI doing the displacing. But Amodei has an explanation for the apparent contradiction: he’s just being honest where others aren’t. “Most of them are unaware that this is about to happen,” he told Axios. “It sounds crazy, and people just don’t believe it.”
The real reason Anthropic CEO Dario Amodei is sounding the alarm
Anthropic is closing in on what could be its final private funding round before an IPO, reportedly targeting a valuation north of $900 billion. That would make it more valuable than OpenAI, which closed a record-breaking round at an $852 billion valuation earlier this year. Anthropic’s annual revenue run rate has reportedly crossed $40 billion—but it’s still burning enormous amounts on compute, research, and talent.To hit a $900 billion valuation, Anthropic needs institutional investors—BlackRock, Fidelity, sovereign wealth funds—to believe Claude isn’t just a productivity tool. It needs them to believe it’s a replacement for human labor at scale. The bigger the addressable market, the more the valuation math works.If AI replaces 50% of white-collar jobs, the total addressable market isn’t enterprise SaaS. It’s the entire global wage bill for knowledge workers. That’s a very different pitch to a pension fund manager deciding whether to write a big check.Ben Goertzel, the scientist who coined the term AGI, put it plainly to Fast Company: “If all jobs are going to be taken over by AI, you better own a piece of that AI.” Amodei probably believes what he’s saying. But Wall Street hears it as an invitation, not a warning. When AI leaders discuss widespread job displacement, they’re also reinforcing the narrative that generative AI will soon take over corporate work tasks at scale—and companies representing roughly a third of US stock market value are already making major bets on that story.The doom-and-gloom circuit also makes for convenient positioning. Anthropic gets to be the responsible actor in the room—the company that warned you and came with solutions—while simultaneously selling the product that is supposedly doing all the disrupting.
What the actual numbers say about AI and jobs
Meanwhile, Anthropic’s own researchers published findings in March 2026 that quietly complicate the narrative their CEO is selling on podcasts. Computer programmers top Anthropic’s AI exposure index at around 75% task coverage. But zoom out to the broader computer and math category, and AI theoretically could handle 94% of tasks. Claude is currently covering just 33%. Legal, office admin, and finance roles show similar gaps between theoretical capability and actual adoption.Crucially, the paper found no systematic rise in unemployment among the most AI-exposed workers. What the data does show is a roughly 14% drop in hiring of 22–25-year-olds into exposed roles since ChatGPT launched—concerning, but a far cry from the bloodbath being described in Amodei’s public appearances. The Yale Budget Lab, separately, found no significant macroeconomic effects from AI on labor through late 2025. “No matter which way you look at the data, at this exact moment, it just doesn’t seem like there’s major macroeconomic effects here,” the lab’s co-founder Martha Gimbel told Fortune.Meta’s ex chief AI scientist and one of AI’s godfathers Yann LeCun made the same point more bluntly. When Amodei’s warnings circulated recently, LeCun posted on X: “Dario is wrong. He knows absolutely nothing about the effects of technological revolutions on the labour market.” He argued that no AI leader—himself included—is qualified to make these calls, and that people should instead “listen to economists who have spent their career studying this.” When someone noted the comments were from an older interview, LeCun was unmoved: “It’s still wrong, destructive, and dangerous.”
Sam Altman , Jensen Huang, and others have quietly changed thier tune
A range of voices in the industry are actively pushing back—including, notably, Sam Altman himself.Altman spent much of early 2026 matching Amodei’s register, warning that AI’s impact on jobs “will begin to be palpable” within a few years. But on May 1, he posted something that sounded almost like a correction: “We want to build tools to augment and elevate people, not entities to replace them.” He followed it up with a note that “jobs doomerism is likely long-term wrong” and that he’s hopeful for a future where people who want to work have “incredibly fulfilling things to do.“hat’s a significant tonal shift from a CEO who was openly musing about AI replacing even himself at the OpenAI CEO role. It also happens to come right as scrutiny over AI-driven layoffs is reaching a crescendo—and as OpenAI eyes its own IPO. When the founder of the most powerful AI company in the world pivots to “augment and elevate,” it’s worth noticing.AWS CEO Matt Garman has been even more direct. At an Amazon event this week, he said the company plans to hire 11,000 software engineering interns in 2026—in line with previous years—and that demand for developers is actually “accelerating.” He acknowledged the role is evolving but pushed back on the replacement narrative: “Being an expert at being able to author a Java code snippet is going to be less valuable,” he said, “but that’s different from the job going away.“Nvidia’s Jensen Huang, who might be expected to benefit most from AI hype, has also been measured—arguing that AI will create new industries and new job categories, including six-figure roles building AI infrastructure. Google CEO Sundar Pichai, whose company now generates 75% of new code with AI, has similarly said the message is adaptation, not apocalypse: “People who learn to adopt and adapt to AI will do better.”A NYT report that interviewed more than 70 software developers across Google, Amazon, Microsoft, and various startups found a different mood on the ground than Amodei’s warnings would suggest. Most developers were, in the reporter’s words, “weirdly jazzed” about AI tools—not afraid of them. Coding, one technologist told the paper, is one of the rare fields where AI takes away the drudgery and leaves the human, creative parts intact. That’s almost the opposite of what Amodei is describing in his public appearances.
Anthropic can’t fill its own roles
The internal contradiction at Anthropic is hard to miss. While Amodei warns about the death of coding jobs, Anthropic has been advertising over 400 engineering roles, some paying up to $405,000 annually. The company’s Head of Growth, Amol Avasare, said recently that product management is “absolutely squeezed” and that Anthropic urgently needs to hire more PMs—even as Claude Code has reportedly tripled engineering output internally.Boris Cherny, who built Claude Code and predicted the “software engineer” title could vanish by end of year, acknowledged that humans still need to review AI-generated code closely. “I don’t think we’re at the point where you can be totally hands-off,” he said on Lenny’s Podcast. He’s stopped writing code by hand himself since November—but he still checks every line Claude produces.So the tool’s own creator says you still can’t fully trust it unsupervised. But the tool’s company CEO is telling the world it’s about to consume all of white-collar work. Somewhere between those two statements is where the actual future of AI and jobs is being worked out—quietly, without much fanfare, and almost certainly slower than the IPO roadshow requires.
