Tech Executives Say AI Cuts Workloads — Employees Report Working Up to 90 Hours a Week

Tech executives publicly claim that artificial intelligence reduces workloads. A BBC News investigation found a gap: employees at the same companies report that their working week often reaches 90 hours, more than double the standard.

What executives promise

Leaders of major tech companies are actively pushing a narrative of AI as a tool that frees people from routine work. In corporate reports and at industry conferences, the same thesis recurs: AI takes over repetitive tasks, from writing code to drafting business emails and analyzing data, letting specialists switch to strategic and creative work.

According to BBC

News, the official position of tech management boils down to this: AI is meant to "free employees for higher-order work."

Executives' key argument is that individual productivity rises while working time stays the same or even falls. This image, less routine and more value, is central to their communication with investors, boards of directors and regulators.

What employees say

Tech employees describe a fundamentally different picture. According to BBC News, industry workers report a working week that often reaches 90 hours, more than double the standard 40.

• A working week of up to 90 hours is the figure cited by the tech employees BBC surveyed
• The rollout of AI tools coincided with rising expectations, not falling workloads
• A specialist who completes a task twice as fast typically gets twice as many tasks
• Work-life balance has worsened despite the declared "liberation from routine"
• The gap between CEO rhetoric and reality is documented by BBC journalists and independent researchers alike

This phenomenon is called the "productivity paradox." Every wave of automation, from the steam engine to spreadsheets, raised productivity but did not reduce total working time: that time filled up with new tasks and higher standards. The AI wave appears to follow the same logic.

Why is the gap so large?

The gap between management's position and employees' experience comes down to different incentives. Executives evaluate AI by corporate metrics: speed of product delivery, lower operating costs, revenue growth. If AI lets a team of 10 do what used to take 15, that is a success on the P&L line, not a reason to cut headcount or hours.

For an employee, the same situation looks fundamentally different: the team has not shrunk, and the bar for expectations has gone up. Tools that were supposed to make work easier have become a justification for higher output quotas. According to several studies, when AI tools raise developer productivity by 20–30%, companies primarily use that gain to expand the volume of tasks, not to reduce the load on people.

What this means

Shorter working hours are not a technological consequence of AI but a management decision. As long as companies treat productivity gains as a chance to take on more tasks for the same money, the promise that AI will "free up people's time" will remain investor rhetoric. A 90-hour working week is a vivid illustration of how far words and reality diverge at tech companies today.