For the past five months, I’ve been reading through studies from some of the best labor economists at Harvard, MIT, NBER, Brookings, Anthropic. I wanted to see whether AI career disruptions were already happening.
The short answer is yes. It’s happening, but almost nothing about it matches the popular “Five things to do NOW!!!” posts.
Unemployment is low, the economy is growing, and AI is disrupting careers. All three of these can be true at the same time.
The overall employment numbers look normal because they’re averages. Underneath them, business and financial occupations have declined roughly 2%. Architecture and engineering are down about 1.3%. Meanwhile, legal occupations rose by 7.6% (despite being AI-exposed).
People are shifting around so it gives the illusion of normalcy.
Anyone who says AI isn’t causing job losses because unemployment is low is not looking at the numbers within the numbers. The disruption is happening by movement within the workforce. There are shifts between occupations, firms, and age groups, even though the current total headcount is stable.
A team of researchers got access to ADP payroll data covering 55 million American workers. They tracked employment changes in occupations with high AI exposure versus low AI exposure, broken down by age group.
They found that workers aged 22–25 in AI-exposed occupations experienced a 6% decline in employment between late 2022 and September 2025. Workers over 30 in those same occupations saw employment grow 6–9%.
In occupations with lower AI exposure, this age gap doesn’t exist. Young workers kept pace with older ones.
AI-exposed work is suppressing entry-level hiring.
The researchers called these young workers “canaries in the coal mine.” Young workers are the first ones to feel the effects. Companies are discovering that AI can handle the tasks they used to hire juniors to do like basic research, first-draft writing, standard coding and analysis.
Back in 1993, I got my first job at a computer lab at Northwestern University. Everyone who worked in that lab was extremely patient and explained to me network connectivity and Unix administration.
It was because of that first job that I was able to springboard into other jobs as a tech support technician, a computer consultant, and on and on from there.
If I was a 23-year-old in this job market, I simply wouldn't have that opportunity. Newer AI systems would perform all the basic tasks that I did to earn a job in that computer lab. AI systems could sort through help desk tickets and answer lower-level support requests. There’s no need for younger employees in an AI-exposed company.
So a 23-year-old wouldn't have that springboard I did, so I'm not sure how they would pick up skills for their career. A career pipeline that I rode for decades has been broken by AI systems.
That’s why it worries me that almost all the advice these young folks see is telling them to solve a different problem. Even if you're a terrific communicator, or a brilliant networker, or filled with emotional intelligence it doesn’t solve the underlying problem of an entry-level job not existing.
As a parent and someone who's worked in technology for over 30 years, I don't want young folks to feel like their career is starting with failure.
So what I can tell you is that you're probably getting a lot of bad advice. If you feel like you're not getting started in your career, it might be because older folks like me have closed off many of the entry-level jobs that boosted our careers.
But it's not just individual employees that are missing out on these opportunities. Organizations that focus on short-term efficiencies are giving up long-term talent cultivation. When nobody comes up through the ranks, the knowledge, culture, and judgment that define your company doesn’t get passed down. It just disappears or goes somewhere else.
It will be a free agent world of hiring, where HR leaders will have to pay more for external talent or promote internally before someone is ready.
By switching out people for AI systems, these organizations are cooking the “seed potatoes” of their future talent pool.
Over the next few months, this newsletter is going to walk through what the research actually shows about AI career disruption. Who's affected, what protects people, and what organizations should be doing about it. I'll keep bringing what I find so you can make better decisions about what comes next.