AI roles jump 20–30% year‑over‑year

The job market story here is not “everyone is hiring again.” It’s almost the opposite. Hiring is still soft in a lot of places, especially across white-collar and tech roles. But inside that weak market, employers keep carving out budget for AI work — and that split is getting easier to see. (hiringlab.org) ### What actually changed? Indeed’s January 22, 2026 labor update put a number on the shift: job postings that mention AI or related skills kept climbing through late 2025, even while overall hiring remained subdued and many occupations were flat or down. By December 2025, the share of postings on Indeed mentioning AI had reached 4.2%, the highest level in that tracker. (hiringlab.org) ### Why does that matter? Because the rest of the labor market is not exactly booming. Total US job postings on Indeed at the end of 2025 were only about 6% above pre-pandemic baseline levels, which is a pretty modest backdrop after several years of churn. In other words, companies are not opening the floodgates. They’re being selective. And one of the clearest things they still want is AI capability. (hiringlab.org) ### Is this really a 20–30% story? Not from the best public data I could find. The stronger number is bigger. Indeed’s data shows the number of postings mentioning AI-related terms surged by more than 130% over the same period that overall postings stayed only modest(hiringlab.org)he cleanest published figure is not 20–30% year over year. It’s a much sharper rise in AI mentions over the last few years. (hiringlab.org) ### Where is the demand showing up? A lot of it is still in technical work, but not only there. Indeed says nearly 45% of data and analytics postings now contain AI-related terms. Marketing is around 15%, and human resources is around 9%. Lightcast shows the same bro(hiringlab.org)s, with 51% of AI job postings now outside IT and computer science. (hiringlab.org) ### So is tech hiring back? Not really. That’s the catch. Indeed’s tech hiring work showed US tech postings were still down 36% from early-2020 levels as of July 2025, and the freeze had already stretched into a third year. But the jobs directly tied to AI — machine learning engineers are the obvious example — were among the few categories still holding up better than the rest. (hiringlab.org) ### Who is doing the hiring? Mostly very large employers. That matters more than it sounds. Indeed found that only about 1 in 20 companies had even one AI-related posting by late 2025, and almost 90% of all AI-related postings came from just 1% of firms. So this is not a broad-based hiring wave washing across the economy. It’s concentrated demand from big companies with money, data, and clear use cases. (hiringlab.org) ### What does that mean for workers? Basically, “AI jobs” increasingly means two different things. One is the obvious bucket — machine learning engineers, data scientists, infrastructure people. The other is ordinary-looking jobs that now ask for AI fluency on top of the old job. That secon(hiringlab.org)org. (hiringlab.org) ### Bottom line? The real signal is selective hiring. Companies are still cautious overall, but they keep spending on roles and skills that promise leverage. That’s why AI hiring can rise in a weak market without contradicting the broader slowdown — it’s not a rebound, it’s a prioritization. (hiringlab.org)