Anthropic CEO Warns AI Will Eliminate Half of Entry-Level Office Jobs in 5 Years
AI to Eliminate Half of Entry-Level Office Jobs in 5 Years

Anthropic CEO Issues Stark Warning: AI Poised to Decimate White-Collar Workforce

In a sobering assessment of artificial intelligence's rapid advancement, Anthropic CEO Dario Amodei has delivered a stark warning that AI is coming for white-collar jobs with unprecedented speed and scale. During a May 2025 interview with Axios, Amodei cautioned that approximately half of all entry-level office positions could vanish within the next five years, potentially driving unemployment rates to alarming levels of 10-20%.

The CEO accused both governments and AI companies of "sugarcoating" the imminent threat, suggesting that most people remain unaware of the approaching disruption. Amodei had previously sounded similar alarms at the World Economic Forum in Davos earlier in 2025, telling world leaders that AI's impact on employment would far exceed what most governments have prepared for.

High-Risk Professions and Internal Contradictions

According to Amodei's analysis, professions in technology, finance, law, and consulting stand at the forefront of this transformation. However, a fascinating contradiction emerges when examining Anthropic's own operations. Despite warning about mass job elimination, the company finds itself desperately seeking to fill traditional tech roles.

Wide Pickt banner — collaborative shopping lists app for Telegram, phone mockup with grocery list

Amol Avasare, Anthropic's Head of Growth, recently revealed that the company is struggling to hire product managers. While Claude Code—the company's AI coding tool—has effectively tripled engineering output, the human professionals directing this work have become overwhelmed. "PM and design is just squeezed. It's absolutely squeezed," Avasare stated. "We need to actually hire a ton more PMs."

This internal hiring challenge at a company warning about widespread job displacement reveals a more nuanced reality about AI's current impact on the workforce.

Research Reveals Gap Between AI Capability and Implementation

Anthropic's own research, conducted in March 2026 by researchers Maxim Massenkoff and Peter McCrory, maps the significant gap between what AI can theoretically accomplish and what it's actually performing in workplaces today. Their study found computer programmers lead in "observed exposure" with a 74.5% score, meaning AI already handles nearly three-quarters of their tracked tasks.

Customer service representatives and data entry workers follow closely behind in exposure levels. However, when examining broader categories, the implementation gap becomes more apparent. In the computer and mathematics field overall, AI could theoretically manage 94% of tasks, but Claude currently covers only 33%.

Similar disparities exist in legal, office administration, and finance roles, where capability significantly outpaces actual adoption. Interestingly, the researchers discovered no systematic increase in unemployment among highly exposed workers, though hiring of 22-25-year-olds into these roles has declined approximately 14% since ChatGPT's launch.

Surprising Profile of Most Vulnerable Workers

The characteristics of workers most exposed to AI displacement present unexpected findings. Anthropic's data reveals these vulnerable professionals earn 47% more than the least exposed group, are 16 percentage points more likely to be female, and nearly four times as likely to hold graduate degrees. This profile includes lawyers, financial analysts, and developers—not the warehouse and manual labor positions many might assume would be most at risk.

Boris Cherny, the engineer who developed Claude Code, predicted in February 2026 that the "software engineer" job title would begin fading this year. However, he envisions the role transforming rather than disappearing entirely, evolving toward spec writing, user research, and product thinking responsibilities.

This evolution circles back to Anthropic's own hiring dilemma. While publicly warning about AI eliminating millions of jobs, the company internally scrambles to fill the very roles that managing AI output continues to make increasingly essential. The contradiction highlights the complex, transitional nature of AI's impact on employment as capabilities advance while human oversight remains crucial.

Pickt after-article banner — collaborative shopping lists app with family illustration