The Jobs AI Is Supposed to Replace Are the Ones Getting Pay Rises
Indeed hiring data shows advertised salaries in the most AI-exposed US occupations up 46 percent since 2021, against 25 percent for the least exposed, even as AI-cited layoffs reached 116,175 through August.
New hiring data from Indeed contains a result that does not fit the usual story. Advertised salaries in the US occupations most exposed to generative AI have risen 46 percent since 2021. Moderately exposed roles rose 41 percent, all posted wages 39 percent, and the least exposed occupations only 25 percent. The gap became visible from 2024 and has widened since. “Exposed” here is a technical term meaning a large share of the job’s tasks are things current AI can do, and the most exposed categories are software development, IT support, data and analytics, marketing, and banking and finance. In other words, the jobs most often named as at risk are the ones where advertised pay is climbing fastest.
This sits alongside a genuinely uncomfortable number from the other direction. Outplacement firm Challenger counted 116,175 layoffs citing AI through August 2026, about 22 percent of all announced US job cuts. Both things are true at once. The reporting also notes the effect is uneven: entry-level roles take the brunt, while senior positions that involve supervising AI systems command the widest premiums. After controlling for which occupations are growing, the data still shows a pay premium of about 5.7 percent for exposed roles since ChatGPT arrived.
What is behind this
The apparent contradiction dissolves once you separate two different questions. “Can AI do parts of this job?” and “Does a person doing this job produce more value than before?” have different answers. When a tool makes a worker substantially more productive, a company can want fewer of them and pay the remaining ones more. That is not a paradox, it is what happened with spreadsheets and bookkeeping, and it is brutal for anyone on the wrong side of it while being invisible in the average.
Two caveats matter. Advertised salaries are what employers post, not what people are paid, and postings skew toward roles that are hard to fill. And the entry-level finding is the one that should worry people most, because the traditional path into these professions ran through exactly the tasks now being automated. If the ladder loses its bottom rungs, the premium at the top is cold comfort.
What this means for you: The practical reading is that “AI-exposed” is not the same as “doomed,” and treating your field as a lost cause because it appears on an exposure list is probably the wrong conclusion. The premium is going to people who direct these tools rather than compete with them, which is a learnable skill and mostly a matter of use rather than study. If you are early in a career in one of these fields, the honest advice is less comforting: the on-ramp is narrower than it was, and building visible work of your own matters more than it used to.
Sources
Alibaba Wants a Ten Trillion Parameter Model, and Its Own Chips to Train It
At its Apsara conference CEO Eddie Wu said Alibaba is planning a Qwen successor of 5 to 10 trillion parameters and unveiled the Zhenwu V900 chip, which the company says triples its predecessor's performance.
