In May 2025, Anthropic CEO Dario Amodei predicted that half of all entry-level white-collar jobs would disappear due to advancements in artificial intelligence. One month later, OpenAI CEO Sam Altman also predicted the end of certain job categories.
However, data from the US Bureau of Labor Statistics shows that the US has continued to add jobs and unemployment has remained at historically low levels since the introduction of AI chatbots in late 2022.
A report titled 'Canaries in the Coal Mine?' from Stanford found a 16% decline in employment for younger workers in jobs exposed to AI, such as customer service and computer programming. US Senator Bernie Sanders cited this research, questioning how millions of people could be thrown onto the street without a plan.
Despite these findings, employment trends in occupations where AI impacts are expected first remain largely stable, according to Erika McEntarfer, a fellow at the Stanford Institute and co-author of the report.
The rise in unemployment for the 20% of workers most exposed to AI since 2022.
Since 2022, the unemployment rate for the 20% of workers most exposed to AI rose by 0.77 percentage points, a figure lower than the 0.85 percentage-point increase seen among the least-exposed workers.
Earlier this year, the unemployment rate for recent graduates reached 5.6%, compared to a national average of 4.2%.
Shifting employer requirements
An employer survey by ZipRecruiter indicates that 57% of employers expect higher productivity from employees due to AI, and 38% have transferred tasks previously performed by inexperienced or young workers to AI.
The survey also found that 31% of employers have increased experience requirements for entry-level positions due to AI. Additionally, 74% of employers consider AI skills a requirement or a strong advantage, and 13% require these skills company-wide rather than just in technical roles.
The clearest trend line is a rising bar rather than a shrinking pool.
Half of the employers polled by ZipRecruiter expect candidates to be advanced or practical AI users upon their first day.
Market turbulence and structural changes
Economics professor Nicholas Bloom of Stanford University described the current job market as experiencing turbulence due to AI. NYU Stern professor Robert Seamans categorized the impact of AI into three areas: jobs created, jobs made obsolete, and jobs changed, noting that 'jobs changed' is the largest category.
At the AI coding platform Bolt.new, a three-person analytics team used an AI agent to save 12 to 13 hours of manual work per week; CEO Eric Simons stated the team's output is equivalent to a 30-to-40-person team.
While some see growth, others express concern. US Representative Greg Casar stated that AI will eliminate jobs for millions of Americans while benefiting a small number of executives and investors. US Representative Alexandria Ocasio-Cortez stated in December 2025 that an overwhelming majority of Americans worry AI will leave them permanently unemployed. US Senator Elizabeth Warren stated in June that the government cannot wait to act if AI could disrupt half the workforce.
Conversely, Taylor Budowich, founder of Innovation Council and former deputy chief of staff in the Trump White House, told Fox News Digital that the AI boom under President Trump will make American workers more employable and productive.
Other perspectives include Paul Osterman, a professor emeritus at MIT, stating that approximately 35% of the US workforce is already considered easily replaceable. Meanwhile, Tim Newman of TechEquity noted that more workers are attempting to negotiate the use of AI within collective bargaining agreements.
