Tech companies promised artificial intelligence would shorten the working week. Instead, staff building and using the technology say they are working longer hours than ever, the BBC has found.
The promise, and the reality
An engineering director at Google predicted four years ago that AI would deliver a four-day working week by 2025. OpenAI later urged other companies to trial a four-day week, arguing AI would soon speed up human labour enough to justify it.
But a former OpenAI technical employee who left the company last year told the BBC that OpenAI never trialled the four-day week itself. Instead, the employee described a culture of frequent “crisis meetings”, weekend work and “super cut-throat” performance reviews that saw colleagues suddenly let go.
“You go in on Saturday or Sunday just to catch up or make sure things aren’t broken,” the person said.
Sprints topping 90 hours
The former employee said they worked at least 70 hours a week at OpenAI. Tech workers told the BBC that “sprints”- intense periods of work before a product launch- at OpenAI and Anthropic can stretch on for many weeks and top 90 hours in a seven-day period. Neither company responded to the BBC’s request for comment.
Being ‘drafted’ at Meta
At Meta, workers said they have been abruptly moved onto urgent AI projects this year without being given a choice- a practice staff called being “drafted”.
“They just move you over,” a former employee said. “You can’t say no – or if you do, you have to quit.”
Meta has since eased that approach, staff said, but many workers had already been reassigned by then. Hours on these teams are often long, with staff working into the night and on weekends, and feeling “on call” even when not working, current and former employees said. A Meta spokesperson declined to comment.
Executives’ promises
Anthropic has said its Claude tool can work unaided for seven hours- a full corporate workday. Meta chief executive Mark Zuckerberg has said AI is starting to “dramatically change” the way the company works, letting fewer employees do more. Meta has since cut one in ten of its staff. Anthropic chief executive Dario Amodei has separately warned that as AI tools become more productive, they could bring wider job losses.
Longer hours beyond AI teams
Workers outside AI development are affected too. Former Google employee Amin Shali left the company in May, saying he often had to work through the night after internal engineering systems failed- a problem he said grew more frequent as Google shifted processing power and memory to AI projects. A Google spokesperson declined to comment. Shali said his sleep and health improved after he left, and that reliance on AI tools “creates a bad culture with excess pressure on engineers”.
The research
An eight-month study by the University of California, Berkeley, followed hundreds of employees at a US tech company using AI tools. It found staff worked at a faster pace, took on a broader range of tasks, and extended work into more hours of the day- partly because of the need to constantly check AI output.
MIT innovation scholar Neil Thompson said any time saved by AI tends to be absorbed by the work of implementing and checking it, or filled with new tasks as employees try to prove their value.
“People assume that 20 per cent less work means four-day weeks,” Thompson said. “But new work emerges.”
‘It seems to be the opposite’
“AI is supposed to be doing so much for us now, so many more people should at least have better health and better sleep,” Shali said. “Instead, it seems to be the opposite.”
