AI Is Reshaping Jobs
New research suggests AI has already prevented a large number of coder jobs from being created, and corporate restructuring tied to automation is keeping tech layoffs elevated. Microsoft Research and other reports frame the effect as uneven: AI compresses routine hiring while creating demand for supervision, integration and higher-skill roles. (nbcnews.com) (microsoft.com)
The first place artificial intelligence hit the job market was not a factory floor but the entry-level coding ladder. A Federal Reserve paper released on March 20, 2026 found that employment in programming-heavy jobs kept growing after ChatGPT arrived in November 2022, but the growth rate slowed sharply enough that researchers estimate about 500,000 coder jobs were never created. (federalreserve.gov) (nbcnews.com) That is a different kind of job loss than a layoff. Instead of companies firing half their engineers overnight, the data points to hiring managers deciding they need fewer junior coders because one experienced developer with an artificial intelligence assistant can cover more routine work. (federalreserve.gov) (microsoft.com) Coding showed up first because it is one of the jobs artificial intelligence is already good at in small pieces. Anthropic’s Economic Index, based on millions of anonymized Claude conversations, found usage concentrated in software development and technical writing, with artificial intelligence helping on specific tasks more often than replacing whole jobs. (anthropic.com) That “specific tasks” point matters because jobs are bundles, not single chores. Anthropic found artificial intelligence use leaned 57 percent toward augmentation, meaning a person and a model working together, versus 43 percent toward automation, meaning the model directly doing the task. (anthropic.com) The squeeze is landing hardest on the bottom rung. ADP Research, summarizing Stanford Digital Economy Lab work from August 2025, said employment for workers ages 22 to 25 in highly artificial-intelligence-exposed jobs fell 6 percent from late 2022 to July 2025, while older workers in those same categories still saw growth. (adpresearch.com) Software development was the clearest example in that dataset. ADP said employment for the youngest software developers in July 2025 was 20 percent below its late-2022 peak, which suggests companies are still hiring coders, just not building the same junior pipeline they used to. (adpresearch.com) Microsoft Research’s April 9, 2026 report describes the new job mix in plain terms: workers are moving from doing every step themselves to guiding, critiquing, and improving what artificial intelligence produces. In other words, the new openings are less “write the first draft of the code” and more “check the draft, connect it to real systems, and catch what the model missed.” (microsoft.com) That helps explain why tech layoffs can stay elevated even while companies keep spending on artificial intelligence. If a business is reorganizing teams around fewer routine producers and more senior reviewers, platform engineers, and product managers, it can cut headcount in one part of the org chart while hiring in another. (microsoft.com) (nbcnews.com) The government’s own labor economists are cautious about calling this a mass extinction event. The Bureau of Labor Statistics wrote in February 2025 that technology-driven job displacement usually unfolds gradually, and that many occupations affected by new tools still grow because the work changes before the occupation disappears. (bls.gov) So the pattern emerging in 2026 is narrower and more unsettling than the old “robots take all jobs” slogan. Artificial intelligence is compressing routine hiring first, especially for early-career knowledge workers, while raising the value of people who can supervise models, stitch outputs into messy real-world systems, and take responsibility when the software is wrong. (federalreserve.gov) (microsoft.com)
