Dario Amodei, chief executive of the AI company Anthropic, said on May 28 that artificial intelligence could wipe out half of all entry-level white-collar jobs within one to five years and push unemployment in the United States to between 10 and 20 percent. He made the remarks in an interview with Axios, and said that governments and companies were largely ignoring the risk.
The warning carries unusual weight because of its source. Anthropic builds the Claude family of large language models and competes directly with OpenAI and Google. Amodei is not an outside critic of the technology; he is one of the people building it, and his company stands to gain from the very capabilities he described.
Which jobs are at risk
Amodei told Axios that the sectors most exposed are technology, finance, law and consulting, with entry-level positions hit first. Those roles involve work that current models already handle reasonably well: reading and summarizing documents, drafting reports and correspondence, basic analysis, and routine customer interaction. As models improve, he argued, companies will stop hiring junior staff for such tasks before they begin cutting existing employees.
White-collar work in this sense is not limited to data entry or answering email. It includes analysts, administrators, customer service agents, and early-career lawyers and accountants. A model that can read a contract, produce a report or handle a customer query in seconds changes the economics of hiring for all of those roles. The open question, in Amodei’s framing, is not whether this is technically possible but how fast it arrives and how societies respond.
A call to policymakers
Amodei paired the forecast with a message to lawmakers. He said that AI companies and governments needed to stop downplaying the risk of mass job displacement and begin preparing for it, according to Axios. He floated ideas such as better public understanding of what the technology can do, help for workers to adapt, and policy measures that could cushion the transition, including taxes on AI companies that he acknowledged would run against his own firm’s interests.
Business Insider and Fortune, which covered the interview, noted that other industry leaders have offered more optimistic views, arguing that AI will create new categories of work as it eliminates old ones. Amodei’s position is that the speed of the current shift is different from past waves of automation, and that the usual adjustment period may not be available this time.
Why the forecast matters
If the prediction holds even partially, the consequences extend well beyond economics. Entry-level jobs are how young, educated workers gain experience and move up. Removing that first rung affects career paths, household incomes and social stability, not just quarterly payrolls. Education systems, vocational training, labor law and social safety nets were all designed around a labor market that assumed such jobs would exist.
Amodei’s forecast is a projection, not a measurement, and labor economists remain divided on how quickly AI will displace work at scale. What his remarks do establish is that at least one leading AI developer is now publicly treating large-scale white-collar displacement as a near-term scenario rather than a distant one, and asking governments to plan for it.
