Leaders warn artificial intelligence threatens half of entry-level white-collar jobs
The tech industry, once a destination for computer science graduates, is rapidly automating its early-career talent pipeline. AI coding tools like GitHub Copilot and Amazon CodeWhisperer can now handle most of the entry-level jobs that junior engineers once did, making it increasingly difficult for young professionals to get into it.
The impact isn’t limited to computer engineers. Earlier this month, Anthropic CEO Dario Amodei offered a grim prediction: 60 minutesreiterated his belief that artificial intelligence could replace half of entry-level white-collar jobs in the next five years and push unemployment to 10% to 20%. “If we look at junior advisors, lawyers, financial professionals — a lot of the things they do, AI models are already pretty good at it,” he said. “It’s hard to imagine there wouldn’t be a significant impact on employment there.”
Other AI and business leaders have echoed his concerns. Geoffrey Hinton, the “godfather of artificial intelligence,” believes that “artificial intelligence will replace everyone for mundane mental tasks.” Salesforce CEO Marc Benioff recently said that artificial intelligence already takes on “nearly half” of the company’s workload. Ford CEO Jim Farley warned that “artificial intelligence will leave a lot of white-collar workers behind,” although he was more optimistic about blue-collar roles. He expects skilled workers in factories to become even more important as they provide “vital human support” in the context of automation.
Several major studies support this trajectory. An August report from Goldman Sachs warned that 6% to 7% of U.S. workers could lose their jobs due to artificial intelligence, especially operations and support staff at large companies. Entry-level workers in the tech industry remain the most affected, but hiring in marketing, consulting, graphic design, office management and call centers has also slowed as companies integrate new automation tools.
A January study by the World Economic Forum found that 41% of 1,000 global employers surveyed expected the adoption of AI to lead to significant job losses. A recent study from Stanford University found that employment among workers aged 22 to 25 in AI-related roles, particularly customer service and clerical tasks, has declined by 13% since 2022. In contrast, older workers in the same field are more likely to find jobs.
Other research suggests that the economic downturn isn’t just caused by artificial intelligence. Negative perceptions of Gen Z workers and broader economic uncertainty are also factors. A study from Hult International Business School found that 89% of employers avoid hiring new college graduates and 39% prefer to use artificial intelligence rather than hire Gen Z employees.
Even so, companies are not cutting jobs uniformly. Some companies are reorganizing—cutting traditional roles while adding new AI-specific roles. JPMorgan CEO Jamie Dimon said that despite the reduction in operational positions, the bank’s overall headcount will continue to grow as the bank hires staff to support artificial intelligence development. Walmart CEO Doug McMillon expressed a similar sentiment at an event in September, saying, “It’s clear that artificial intelligence is going to change almost every job.” He said Walmart would eliminate and create jobs with the goal of helping workers “get through this.”
Ironically, while AI tools increase the speed of code generation, the durability of code still depends on human expertise. Code written for AI can be clumsy or brittle, and when complex systems fail, engineers must know how to fix and improve them. Amoudi warned that experienced engineers will still be needed to oversee AI systems, but even these roles may be reduced. “Ultimately, all these islands will be taken away by artificial intelligence systems,” he said in 60 minutes. “Then we’ll eventually get to the point where AI can do everything a human can do. I think that’s going to happen in every industry.”
