Artificial intelligence may be making workers more productive - but it is also making it far harder for young people to land their first job.
New analysis from a Federal Reserve economist suggests experienced professionals could actually see their wages rise as AI boosts their productivity.
But entry-level workers - many of them in Gen Z - may increasingly struggle to get hired at all.
That stark warning comes from J Scott Davis, an assistant vice president at the Federal Reserve Bank of Dallas, who examined employment and wage trends since the launch of ChatGPT in late 2022.
His findings suggest AI is increasingly replacing the routine tasks traditionally handed to younger workers while making experienced employees more valuable.
Overall employment in the US has continued to grow since AI tools exploded into the mainstream, increasing roughly 2.5 percent since late 2022. But the picture looks different in industries most exposed to AI.
Employment at software and IT services companies has fallen about 5 percent, and more broadly jobs have declined around 1 percent across industries most exposed to artificial intelligence.
Twitter co-founder Jack Dorsey's new payments company Block recently cut 4,000 jobs, nearly half of its workforce, as the company restructures around artificial intelligence.
Artificial intelligence may be making workers more productive, but it is also making the job market more difficult for young people trying to get their foot in the door
Amazon has also been slashing corporate roles as it leans more heavily on automation. In January the company announced 16,000 job cuts worldwide, just three months after eliminating roughly 14,000 positions.
Chief executive Andy Jassy has repeatedly warned that AI will shrink the company's workforce over time.
Crucially, those declines appear to be falling most heavily on the youngest workers.
Research shows the drop in employment in AI-exposed sectors is particularly pronounced among workers under the age of 25, while employment totals for older workers have not declined.
In many cases the issue is not layoffs but hiring. Companies are simply bringing in fewer new workers.
'The job market is getting very tough for new graduates in AI-exposed fields,' Davis notes in his analysis.
The shift largely comes down to the type of knowledge involved in many entry-level jobs.
Davis distinguishes between codified knowledge, which refers to information that can be written down, learned in textbooks or taught through formal instruction, and tacit knowledge, which is the type of understanding gained through years of experience on the job.
AI systems are increasingly capable of replicating codified knowledge. They can summarize information, draft reports, write code and produce analysis that once required a junior employee.
But they struggle to replicate the judgment and intuition that come from real-world experience.
'If AI can replicate codified knowledge but not tacit knowledge,' Davis writes, 'AI will automate jobs requiring codifiable knowledge but complement jobs demanding experiential tacit knowledge.'
That dynamic means tasks that used to be the domain of junior workers are among the easiest for AI systems to perform.
At the same time, experienced workers may actually benefit from the technology.
Even though employment growth has lagged in some AI-exposed sectors, wages in those industries have risen faster than the national average.
Average weekly wages across the US have increased about 7.5 percent since late 2022, while wages in computer systems design have climbed nearly 17 percent.
The pattern suggests AI may be helping experienced workers become more productive rather than replacing them.
'AI can substitute for entry-level workers... while at the same time complement experienced workers,' Davis explains.
In practice that means seasoned professionals can use AI to handle routine work and spend more time focusing on tasks that require judgment, strategic thinking and client relationships.
The findings raise an uncomfortable question for young job seekers. For decades many white-collar careers have followed a predictable path.
Graduates start in entry-level roles performing routine work, gradually learn the deeper skills of the profession, and eventually move into more experienced positions where judgment and expertise matter most.
But if artificial intelligence can handle many of those routine tasks, companies may decide they simply do not need as many beginners.
'Firms are going to find that AI is making this method of employee development cost-ineffective, at least in the short run,' Davis warns.
That could leave many younger workers struggling to find their first opportunity in fields increasingly shaped by artificial intelligence.
In the long run, however, companies may face a problem of their own. Without entry-level hires today, there will be fewer experienced professionals tomorrow.
'Leaving new employees off the job ladder is not sustainable in the long run,' Davis says.
Instead, businesses may ultimately need to rethink how young workers gain experience and develop the tacit knowledge that artificial intelligence still cannot replicate.