AI could create these new jobs despite gloomy forecasts, experts say
The rapid adoption of artificial intelligence tools has unleashed doomsday predictions of a broken career ladder and massive job losses. While AI will likely upend some positions, analysts told ABC News, the technology could deliver job opportunities for many workers.
Companies may call upon employees to manage or quality-check AI tools, while other new roles could demand complex or creative thinking as humans complement automated tasks, some analysts said. But, they added, questions still abound regarding jobs to be created by a possible AI-driven overhaul of the U.S. workplace.
"It's not like AI is this tidal wave where we have no control – there are places where we do have control," said Harry Holzer, a professor of public policy at Georgetown University and a former chief economist at the U.S. Department of Labor.
Forecasts of job disruption wrought by AI vary dramatically.
Dario Amodei, CEO of Anthropic, which created an AI model called Claude, told Axios in May that the technology could cut U.S. entry-level jobs by half within five years.
By contrast, the World Economic Forum surveyed 1,000 large companies worldwide, which identified AI as the top driver of potential job gains by 2030. Overall, the technology would help create 170 million jobs across the globe over the next five years, far exceeding 92 million jobs lost, the survey found.
Going back centuries, technological breakthroughs have typically created more jobs than they've destroyed, though AI presents a novel test, Ethan Mollick, a business professor at the University of Pennsylvania and the author of "Co-Intelligence: Living and Working With A.I," told ABC News.
"Historically, technology has resulted in job creation that has offset job loss. Every time that happens we worry, 'It will be different this time around.' It may be different this time around – AI is a very different technology. But we don't know the shape of that."
While observers await the full effects of new technology, AI-themed positions are already popping up at firms across a range of sectors, Chris Martin, lead researcher at job-posting site Glassdoor, told ABC News.
The share of job listings taken up by AI-specific roles more than doubled from 2023 to 2024, according to Glassdoor data shared with ABC News. So far this year, the share of such positions has climbed another 56% compared to the same period a year earlier, the company said.
AI jobs break down into two categories, Martin said. A relatively large set of positions is made up of previously available roles retrofitted for AI-related tasks, such as software engineers or attorneys who specialize in AI. A smaller slice of jobs owes its skills and tasks solely to AI, including a fast-growing set of AI training roles that help optimize the technology through individual expertise in foreign language or another subject.
AI training roles, which typically operate on a freelance basis, grew more than four times in 2024 compared to the year prior; and they've continued to grow in 2025, Glassdoor said.
Some AI-specific jobs have waned, however. Listings have faded for "prompt engineers," AI operators who craft queries to yield useful AI responses, Martin said. "That has mostly gone away," Martin added.
Looking ahead, the path for new AI roles over the coming years remains murky, analysts said. Some analysts said a new category of jobs could require humans to evaluate the quality and authenticity of AI output, while others said AI may become proficient enough to negate many of those roles.
Some analysts touted the possibility that AI tools could do away with generalist roles and yield a boom in specialty positions, such as transitioning a primary care doctor to a role narrowly focused on diagnosis. While others said the rise of AI expertise could threaten jobs even in some specialized professions.
The ultimate capability of AI tools could help determine the scale and type of job opportunities, analysts said. "The question in some ways is: What happens next with these systems," Mollick said.
Predictions of newly created jobs pose a challenge for forecasters, since the positions will emerge in a yet-to-be realized economy transformed by AI, David Autor, a professor at the Massachusetts Institute of Technology who specializes in technological change and the labor force.
"We're not good at predicting what the new work will be; we're good at predicting how current work will change," Autor said.
The murky path forward should engender caution as workers weigh how best to adapt to the changes promised by AI, Mollick said.
"The worst thing you could do right now is make a complex career decision based on what AI is doing today, because we just don't know," Mollick said.
