While tech executives claim AI will reduce work hours, employees at major firms report grueling schedules often exceeding 90 hours per week.

Executives at major technology companies have long argued that artificial intelligence will eventually reduce human labor requirements. However, employees at firms like OpenAI, Anthropic, and Meta report that the reality involves increasingly long hours, weekend work, and intense pressure, contradicting the narrative that AI will lead to shorter work weeks.

Contradictory Workplace Realities

Despite OpenAI publicly suggesting that companies should test four-day work weeks, a former technical employee stated the company never implemented such a policy. Instead, staff described a culture of crisis meetings and weekend work. Similarly, workers at Anthropic and OpenAI reported that during intensive project periods known as 'sprints,' work weeks can exceed 90 hours.

Pressure at Meta and Google

At Meta, employees reported being 'drafted' into urgent AI projects without the option to refuse, leading to long nights and constant on-call expectations. Former Google employee Amin Shali reported leaving the company in May, citing increased pressure and night work caused by the firm shifting resources toward AI projects, which he claimed negatively impacted his health.

Research on AI Workload Impact

Research from UC Berkeley tracking hundreds of tech workers over eight months found that AI tools led employees to work at a faster pace and extend their hours. MIT innovation scholar Neil Thompson noted that any time saved by AI is often absorbed by the need to implement, test, and verify the output of these tools, or is simply filled with additional tasks.