October 7, 2026, (Inside AI) — The rapid adoption of artificial intelligence has not triggered mass job losses but is quietly reshaping tasks and hitting early-career workers hardest, according to Nela Richardson, chief economist at payroll processor ADP. In a wide-ranging interview, Richardson said the data shows AI's impact is "very minute" and best viewed through a microscope rather than a telescope. "Right now, it's better to use a microscope than a telescope, because the changes that we're seeing are actually at the task level," she said. "You're not seeing them in the overall productivity numbers yet."
Richardson's comments come as economists and policymakers debate AI's effect on employment. ADP, which processes payrolls for one in six U.S. workers, has partnered with Stanford University to dissect its data by occupation and AI exposure. The findings challenge the assumption that AI would replace expensive older workers with cheaper, tech-savvy young ones. Instead, AI appears to augment experience, boosting employment for older workers while reducing it for entry-level roles in highly exposed fields like software development and customer service.
Early-Career Workers Face The First Wave
Since the rollout of ChatGPT in late 2022, ADP data shows a 4.4% year-over-year decline in employment for young people in AI-exposed fields as of August, compared to a 2% decline in non-exposed fields. "The rate is faster for AI exposed fields," Richardson said. "But here's why it's counterintuitive... AI seems to reinforce knowledge and expertise." Tasks that are automatable, such as routine customer service or basic coding, are typically done by early-career workers. Older workers, by contrast, are seeing employment rise because AI enhances their judgment and experience.
This pattern is not uniform. In healthcare, a non-AI-exposed sector, employment for young people is rising. "If you look at non-AI exposed fields, like people in the healthcare profession, home health, nursing aids, you're actually seeing employment rise for young people," Richardson noted. The divergence suggests AI's impact is occupational and task-specific, not a broad jobs apocalypse.
Wages tell a similar story. ADP Research finds that occupations tied to machine learning and AI command a significant wage premium over other digital jobs. "We are also seeing in our own research at ADP Research that if you look under the hood of pay in that digital job category, occupations tied to ML, machine learning, and AI are actually seeing a huge wage premium versus other digital jobs," Richardson said. Even in customer service, AI tools can boost performance for low-performing reps, acting as an enhancement rather than a replacement.
Productivity Puzzle And The Aging Workforce
Despite AI's task-level effects, macroeconomic productivity numbers have yet to show a boost. "The first place you should look is productivity numbers, and you sure ain't seeing it there yet," Richardson said. She attributes this to a lack of investment in human capital. "We don't have 30 years to skill up young people to take advantage of AI. We have do it now," she warned. "Until we start investing in people, we're not gonna get the promise of technology."
Meanwhile, an even larger force is reshaping the U.S. labor market: the retirement of the baby boomer generation. "If I could think of two A's that are most impacting the labor market right now, AI or aging, it would be aging hands down," Richardson said. Every day, 10,000 Americans turn 65. This demographic shift is creating labor shortages in skilled trades, especially construction tied to the AI data center buildout. Wage premiums for job switchers in construction are hitting almost 14%, driven by demand and limited immigration.
"We've seen that split... in the last couple years," Richardson said, referring to the divergence between white-collar knowledge jobs and blue-collar trades. "This premium is a signal that this is a much tighter labor market in the good sector than it is often." The healthcare sector, meanwhile, has generated three out of every four net new jobs in recent years, fueled by an aging population that is choosing to age in place.
Inequality remains a concern. Research with the University of Chicago shows that from 2021 to 2024, 43% of workers saw real wage declines, with an average drop of 8%. AI's enhancement of high-skilled salaries could widen the gap, especially in a low-hire, low-fire labor market where job switching, the traditional route to raises, is less frequent. "All of this stuff... is why economics is the dismal science," Richardson said. "We look at both sides of the equation."
Looking ahead, Richardson urges employers to upskill workers to high-value tasks and young people to lean into networking and people skills. "In a time of tech, it's about going back to people and relationships," she said. As for the doomsday scenarios, she remains hopeful. "If we use the tools right, we can reduce the impact, but yeah, the transition is messy. People get hurt along the way. What we want to have is the ability to take the people, upskill them into jobs that not only give them higher standards of pay, but durable skills."
