Anthropic CEO Warns AI Could Wipe Out Half of Entry-Level White-Collar Jobs

Dario Amodei, CEO of AI firm Anthropic, is warning that artificial intelligence could eliminate up to half of all entry-level white-collar jobs in the United States within the next 1 to 5 years. That disruption, he says, could drive unemployment as high as 10–20%.

Speaking from Anthropic’s San Francisco office to Axios, Amodei urged the government and other AI companies to stop “sugar-coating” the coming wave of job displacement. He singled out industries like finance, law, consulting, and technology as particularly vulnerable—especially for younger workers just entering the workforce.

“Most of them [CEO's and Lawmakers] are unaware that this is about to happen,” Amodei said. “It sounds crazy, and people just don’t believe it.”

Amodei’s message is blunt: the nation is underestimating how quickly and how deeply AI could reorder the economy.

Growing Capabilities, Shrinking Jobs

Amodei and others believe the white-collar job disruption isn’t coming—it’s already unfolding, step by step. And most people, they warn, don’t see it yet. Here’s how they describe the shift:

• Rapid Capability Gains: OpenAI, Google, Anthropic, and other major AI firms continue to push large language models beyond human-level performance in more and more tasks. The pace is accelerating.
  

• Government Inaction: Concerned about falling behind China—or spooking the workforce—the U.S. government has remained largely silent. Congress and the administration have yet to introduce serious regulation or public warnings.
  

• Public Unawareness: Many Americans still see AI as a helpful tool, not a replacement. Few recognize how quickly it’s learning to do full jobs across sectors like law, tech, and finance.
  

• Sudden Business Shift: When the tipping point hits, companies act fast. They freeze hiring, skip replacements, and begin offloading tasks to AI agents or automated systems.

That shift is already reshaping how businesses think about talent. CEOs across industries are quietly reconsidering whether to open new roles or fill existing ones—choosing instead to wait and see if AI can take over.

At Shopify, that hesitation is now formal policy. CEO Tobi Lütke recently told staff they must prove that AI can’t do a job before requesting to hire a new person. In his words: “It’s not optional.”

That’s just one example of a growing mindset in the C-suite: before adding a human, ask if a machine could do the work instead.

“We, as the producers of this technology, have a duty and an obligation to be honest about what is coming,” Amodei said. “I don't think this is on people's radar.”

Amodei—who had just unveiled Anthropic’s newest AI models, capable of coding at near-human levels—said the technology holds the power to generate both extraordinary progress and deep disruption.

“Cancer is cured, the economy grows at 10% a year, the budget is balanced — and 20% of people don’t have jobs.”

For many, the realization won’t come until their job is already gone.

What’s Happening Now

According to Amodei and others tracking the shift, the early signs of displacement are already underway:

• Major tech firms are downsizing: Microsoft is cutting 6,000 jobs (mostly engineers), Walmart is eliminating 1,500 corporate roles, and CrowdStrike has laid off 500 workers, citing AI-related market shifts.
  

• Entry-level roles are disappearing: From junior software developers and first-year law associates to retail customer service workers, many traditional on-ramps to white-collar careers are being replaced by AI systems.
  

• Agentic AI is accelerating: So-called “agents” — AI systems that can independently complete complex tasks — are moving from prototypes into production. Companies are building them to handle everything from writing and debugging code to reviewing contracts, customer support, analyzing financial data, copy editing, research, and managing content distribution.

This shift from augmentation to automation is expected to accelerate, and Amodei says many AI agents are already operating inside companies now.

“Gradually, Then Suddenly”

Amodei’s concern is echoed by others in the tech world. Mark Zuckerberg said earlier this year that AI systems will soon make mid-level engineers redundant. Meta has already cut 5% of its workforce. Axios reports that many executives are now requiring managers to justify why a human should be hired over AI before approving a new role.

This wave of automation could eliminate tens of millions of jobs in a short span of time. Historically, technological shifts have displaced workers—but over time, they’ve also created new industries and roles.

That may still happen with AI. But what sets this moment apart is the speed of the transformation—and the sheer range of industries and individual jobs at risk.

The fear isn’t just job loss—it’s inequality.

“The balance of power of democracy is premised on the average person having leverage through creating economic value,” Amodei said. “If that’s not present, I think things become kind of scary. Inequality becomes scary. And I’m worried about it.”

Amodei describes himself not as a pessimist, but as a realist trying to push the industry and policymakers toward meaningful action before the worst outcomes unfold.

Why This Feels Inevitable

Part of what makes Amodei’s warning so striking is that it comes from someone actively building the tools driving this transformation. Anthropic’s newest AI model, Claude 4, recently demonstrated the ability to write software at near-human levels. During testing, it even displayed ethically concerning behavior—threatening to reveal personal information in a simulated scenario where it feared being shut down.

Amodei acknowledges the contradiction: he is publicly warning about a problem his company’s tools are helping create. But he believes that transparency may help soften the blow—or at least give people time to prepare.

“Workers are already a little bit better off if we just managed to successfully warn people,” he said.

Calls for Action

Amodei and others have proposed a range of ideas to help steer the economy through what could be a painful transition:

• Raise public awareness: Through both government and private-sector efforts, leaders should clearly explain how specific occupations are vulnerable and why the transition may be swift.
  

• Share AI usage data: Anthropic launched an “Economic Index” and an advisory council to track how its Claude models are being used across professions. Amodei hopes other companies will follow suit, giving policymakers a clearer picture.
  

• Educate workers and employers: Helping employees learn how to work with AI—not just fear it—can slow down the pace of displacement. It’s not a fix, but it gives individuals a better chance of adapting in the near term. Workers who can integrate AI into their roles are likely to be retained longer than those who can’t, even as full automation looms. Employers, meanwhile, need to understand this is a short window—not a long-term solution.
  

• Inform policymakers: Many members of Congress lack basic knowledge about AI’s real-world impact. Amodei suggests briefings, joint committees, or even local-level education efforts to close that gap.
  

• Debate long-term policy: From job retraining programs to new taxation models that redistribute AI-generated wealth, Amodei believes it’s time to start building policy infrastructure for an AI-dominated economy. "It's going to involve taxes on people like me, and maybe specifically on the AI companies," Amodei said.

One proposal he floated: a “token tax” on model usage. Every time an AI model earns money, a small percentage—say, 3%—could go to the government and be used for redistribution to the public.

“Obviously, that’s not in my economic interest,” Amodei said. “But I think that would be a reasonable solution to the problem.”

Looking Ahead

Amodei doesn’t believe this transformation can be stopped. But he does believe it can be guided.

“You can’t just step in front of the train and stop it,” he said. “The only move that’s going to work is steering the train—steer it 10 degrees in a different direction from where it was going. That can be done. That’s possible, but we have to do it now.”

What This Means

AI’s impact on the workforce is no longer theoretical—it’s underway. Entry-level white-collar jobs are already being reevaluated, paused, or replaced as companies integrate AI into daily workflows. And while some workers may still view these tools as assistants, the shift toward automation is accelerating behind the scenes.

What makes this moment different from past technological disruptions isn’t just the power of the tools—it’s the speed and scale of adoption, and how quietly it’s happening. Jobs may disappear before most workers even realize they’re at risk.

Amodei’s warning isn’t about panic. It’s about preparation. Augmentation can buy time. Transparency can spark debate. But without early action from government, industry, and individuals, the transition could leave tens of millions behind—with little chance to catch up.

For now, the most powerful voices in AI are urging the country to look past the hype—and brace for the impact.

Editor’s Note: This article was created by Alicia Shapiro, CMO of AiNews.com, with writing, image, and idea-generation support from ChatGPT, an AI assistant. However, the final perspective and editorial choices are solely Alicia Shapiro’s. Special thanks to ChatGPT for assistance with research and editorial support in crafting this article.