Studies show AI is reducing some repetitive jobs and reshaping others

In short: New research suggests AI is shrinking demand for some routine, repetitive jobs, while changing many other roles instead of fully replacing them.

What's going on

AI systems tend to do best at structured work that follows clear rules, like sorting data, summarizing information, and handling routine requests. You can think of it like a very fast assistant for tasks that have a checklist. That makes jobs with lots of repeatable steps more exposed to cutbacks or big changes.

Several recent data points show this shift. Job postings for repetitive roles fell 13% after ChatGPT launched in 2022, while postings for analytical, technical, or creative roles rose 20% through March 2025, with finance and tech seeing some of the biggest drops. A 2025 survey of 2,000 US adults also found 20% of full-time workers said AI already automates parts of their job, compared with 15% who said AI added new tasks for them.

Other research suggests the impact is uneven, not uniform. In jobs with high AI exposure, employment for young workers ages 22 to 25 fell 6% from late 2022 to July 2025. MIT research covering 2010 to 2023 found that when AI can handle most tasks in a role, those roles shrink, but jobs where AI only handles some tasks can grow because people spend more time on human strengths like coming up with ideas and checking work.

What to watch

The next question is whether “replacement” keeps outpacing “helping” as newer generative AI tools spread. Watch for continued declines in entry-level office roles, and rising demand for workers who can use AI tools safely and effectively, with clear human oversight.

Source: NYTimes