Big Tech leaders have pitched AI as a route to shorter workweeks, but employees at the companies building those tools report the opposite: longer hours and relentless “sprints.” Current and former staff at OpenAI, Anthropic and Meta describe 70–90-hour weeks, weekend “crisis meetings,” and abrupt reassignments to urgent AI projects. Some blame resource shifts to AI initiatives for failures that require overnight fixes. Research from UC Berkeley finds AI can expand workloads by speeding the pace of work, widening task scope and extending hours, while MIT’s Neil Thompson says any time saved is often consumed by implementing and checking AI outputs. Europe caps weekly hours at 48; the U.S. has no such limit—leaving tech workers to shoulder the productivity push even as executives tout a future of fewer hours.
Related articles:
Generative AI at Work: Evidence from a Customer Support Tool in a Large Software Firm
Microsoft Work Trend Index: How AI is reshaping work
