Companies that invest heavily in AI have grown headcount by 10.2% in the two years after adoption, and entry-level headcount has grown by about 12%, according to a working paper from fintech Ramp and Revelio Labs covering more than 21,000 U.S. firms. That research, based on Ramp’s customer spending data, anchored a panel on the geopolitics of AI at the Deel AI Policy Summit, held virtually on July 9.
The finding set the tone for a conversation between Taylor Stockton, chief innovation officer at the U.S. Department of Labor, and Kanishka Narayan, U.K. minister for AI and online safety, moderated by Axios senior tech policy reporter Maria Curi. Both described evidence showing jobs being reshaped by AI, rather than widespread net job destruction, and both described a rethinking of national training strategy built on apprenticeships and broad-based AI literacy.
Read more: Trump’s Labor Secretary pick could reshape AI enforcement
The data behind the layoff headlines
Curi opened the workforce discussion by pointing out that U.S.-based employers announced 97,006 job cuts in May, the highest total for that month since the pandemic layoffs of 2020, according to Challenger, Gray & Christmas. The outplacement firm’s report found AI led all reasons for job cuts for the third consecutive month, accounting for 40% of May’s announced reductions.
Stockton pushed back on reading those headlines as an AI displacement story. He pointed to research attributing much of the tech-sector cutting to post-pandemic overhiring, the low interest rate environment of that period and remote work arrangements that underperformed for some roles, particularly entry-level white-collar jobs.
“This notion and narrative that there’s going to be widespread job loss across the economy is simply not what the data is sharing,” Stockton said.
The Ramp research found, however, that hiring gains appear only at firms in the top third of per-employee AI spending, and headcount growth generally lagged adoption, showing up several months later, often in the six‑ to 12‑month range. Companies with lighter AI investment saw no statistically significant change.
Narayan said U.K. data points in the same direction. “We’re finding effectively no net aggregate impacts of AI adoption in terms of job losses,” he said. He added that British firms that adopt AI are growing revenue, hiring more, and entry-level graduate jobs are holding steady. Ramp’s data also shows that entry-level workers made up a larger share of heavy adopters’ workforces by the end of the two-year study.
To close the data gap that fuels speculation, the Department of Labor is launching an AI Workforce Hub, a data-sharing partnership with roughly 40 private sector companies including Ramp, AI labs, staffing firms and job boards.
In an exclusive interview with HR Executive at HR Tech 2025, Keith Sonderling, acting U.S. Secretary of Labor, described how this AI Workforce Hub will provide centralized access to upskilling strategies, labor market data on AI’s impact across industries and training programs designed to help close workforce skills gaps.
Read more: The biggest workforce shortage in U.S. history is being overshadowed by AI
A fresh look at education
Stockton shared the department’s America’s Talent Strategy, which he said begins with the admission that decades of federal workforce investment have underdelivered. “It has failed too many American workers. It has failed too many American businesses who are looking for skilled talent,” he said. The administration’s answer is a major expansion of registered apprenticeships into technology fields, with reduced red tape in registration and compliance and new investment in industry intermediaries.
Stockton also highlighted the department’s Make America AI Ready initiative, an AI 101 course delivered entirely by text message in 10-minute daily lessons over seven days. He said more than 100,000 Americans have enrolled since its launch earlier this year by texting READY to 20202.
The U.K. has made a notable symbolic move, dropping its long-standing cross-party target of sending 50% of young people to college. Narayan said future education policy will be coupled far more tightly to employer demand through technical and vocational routes, backed by a goal of training 10 million workers in basic AI skills by 2030. More than 1.7 million courses have been delivered in the past 12 months.
The U.K. is also in the process of redesigning Jobcentre Plus, its national employment service, around large language models that scan resumes, evaluate skills and match citizens to local vacancies. “Nothing is more right for LLM-based support than to bring frontier technology to the dignity of workers across the U.K.,” Narayan said.
Asked what he would tell a worker worried about losing their job, Narayan offered the same advice he gives policymakers: “The best insurance policy, to make sure you get good policy and you make sure you have a decent career path, is to dive in and test these tools.”
*On June 29, 2026, President Trump nominated Keith Sonderling to serve as Secretary of Labor permanently, pending Senate confirmation.
Credit: Source link
