Artificial-intelligence “agents” are starting to bite into payrolls, according to Goldman Sachs economist Elsie Peng. In a new note, Peng estimates AI substitution reduced average monthly U.S. payroll growth by about 25,000 jobs over the past year and nudged the unemployment rate up 0.16 percentage point. Offsetting gains from AI “augmentation” added roughly 9,000 jobs a month and trimmed unemployment by 0.06 point, leaving a net drag of 16,000 jobs monthly and a 0.1-point rise in joblessness. The pressure is falling heaviest on less-experienced workers and in industries with high AI-substitution potential since the debut of ChatGPT in 2022. Companies from Block and Amazon to Oracle and Meta have announced sizable cuts—Block’s move reduced headcount by about 40%—as employers in AI-exposed sectors report a 4% net reduction in roles, a Morgan Stanley survey found. Executives say the technology is rapidly automating white-collar tasks; some, including Circle CEO Jeremy Allaire, argue workers who embrace “agentic” tools will see their impact grow even as routine roles disappear.
Related article:
— Generative AI at Work: The Impact of Large Language Models on Call Center Productivity
