AI-attributed US job cuts reached 116,175 in the January-August 2026 period, more than double the 54,836 recorded for all of 2025, with AI's share of total announced job cuts climbing from 7% in January to nearly 40% by May. The acceleration within a single year is the more striking part of the data, more than the raw totals themselves.
A closer read of the role-level data complicates the simple narrative that AI is broadly replacing workers: the roles most exposed are specifically the ones with high overlap to current AI capabilities, programming, customer service, and content work, while demand has held up or grown for machine learning infrastructure, AI safety, and applied research roles. That's consistent with substitution concentrated in a specific band of tasks rather than displacement spread evenly across the labor market.