Tech companies have spent years promising that AI will help people accomplish more in less time. Inside several companies leading that race, some employees are describing an awkward consequence. More productive tools haven’t necessarily produced shorter working days.
Workers at OpenAI, Anthropic, Meta, and Google told the BBC about punishing schedules and AI projects consuming nights and weekends. The twist is that their employers increasingly have evidence that AI really can help people get more done.
AI really can help workers do more
OpenAI says Codex has become the primary AI work tool across every department, including Legal, Finance, and Recruiting. As its capabilities improved, employees increasingly handed it longer and more complex jobs, while workers also used it to tackle tasks outside their usual roles.
Anthropic’s internal numbers point in the same direction. More than 80% of code merged into its codebase was authored by Claude as of May, while the typical engineer merged eight times as much code per day during the second quarter of 2026 as in 2024. Anthropic cautions that code volume overstates the true productivity gain, but says the increase still indicates substantial acceleration.
Employees aren’t getting that time back
A former OpenAI technical employee told the BBC they worked roughly 70 hours a week, sometimes using weekends to catch up. Workers at OpenAI and Anthropic said intense release sprints could exceed 90 hours in a seven-day period.
At Meta, one current and one former employee described workers being abruptly moved onto urgent AI projects, with some assignments stretching into nights and weekends. Former Google employee Amin Shali said AI priorities affected him more indirectly. He linked overnight work to internal engineering problems as resources were redirected toward AI projects, and said his sleep and overall health improved after leaving Google.
Productivity gains can become workload gains
In-progress UC Berkeley research offers a possible explanation. During an eight-month study at a 200-person tech company, researchers found employees using generative AI worked faster and expanded what they attempted. Work also crept into periods that previously acted as breaks, often without managers explicitly asking employees to work longer.
OpenAI itself has proposed another outcome for those efficiency gains. Its Industrial Policy for the Intelligence Age suggests incentivizing 32-hour or four-day workweek trials when AI reduces routine workloads and operating costs, with reclaimed hours potentially becoming permanently shorter weeks or paid time off.
That may be the more important AI productivity question. If these tools really can give companies hours of human labor back, who gets to keep those hours?
