Diego Amodei, CEO of Anthropic, warns that artificial intelligence could eliminate half of all entry-level office jobs and raise unemployment rates to 10–20% over the next five years, according to Axios.

Ironically, his statements came after the company presented its new generations of AI models, Claude Opus 4 and Sonnet 4, which can work longer on tasks and execute them with greater accuracy. This potentially heightens the risk that Amodei warns about.

Amodei urges the U.S. government and companies developing AI to stop downplaying the threat: he states that lawyers, consultants, financiers, and newcomers in IT will face massive layoffs as AI systems increasingly perform routine tasks at or above human levels. "Most people are not even aware that this is about to happen," he noted, suggesting official briefings for lawmakers, requiring companies to disclose AI usage data across various industries, and introducing a tax on AI profits to fund retraining programs for displaced workers.

According to Amodei, AI has the potential to aid in cancer treatment, boost the economy, and balance budgets; however, without proactive measures to preserve jobs and distribute benefits, millions will find themselves unemployed, and inequality will only deepen.

Already, large companies are replacing workers with AI assistants: Mark Zuckerberg predicts that mid-level programmers will begin losing their jobs this year. In January, he stated on Joe Rogan's podcast: "By 2025, we at Meta, along with other companies, will have AI that can effectively perform the functions of a mid-level engineer, capable of writing code." He mentioned that this will ultimately reduce the need for humans in such roles. Shortly after, Meta announced plans to cut 5% of its workforce. Amodei added, "As soon as businesses see the profit from replacing humans with AI, they will do it on a large scale."

To prevent this scenario, Amodei suggests:

• Publishing real-time data on AI usage by profession;
• Encouraging the augmentation of human work rather than its complete replacement;
• Introducing a "token tax" on AI transactions to redistribute profits.

"We can't just stand in front of the train and stop it, but we can slightly change its direction. We need to do this now," concludes the CEO of Anthropic.