For all the fantastical claims of AI rendering millions of white-collar jobs irrelevant or obsolete and unlocking tremendous economic value and productivity, its real-world effect in the workplace so far has been mild at best, it seems. That’s according to OpenAI’s new report on enterprise AI adoption, which reveals that AI is saving workers time and effort, but not by much, with only 40 to 60 minutes saved per day.
OpenAI surveyed 9,000 workers across more than 100 companies and used real-world usage data from enterprise customers, and the key takeaway is that businesses are rapidly adopting AI and “experiencing measurable productivity and business impact”.
AI's impact on the workplace is real, but still modest
75% of workers across departments reported that either their speed or the quality of their work improved as a result of using AI. Power users were able to save even more time, up to 10 hours a week. The report also notes that AI is not only speeding up existing workflows, it’s also facilitating performing tasks that workers couldn’t perform previously.
On paper, AI saving an hour a day on average is commendable. But it doesn’t quite begin to measure up to the kind of future we were promised when ChatGPT exploded onto the scene three years ago.
Back in May 2025, Anthropic CEO Dario Amodei warned that AI is a train that can’t be stopped and that the technology could end up wiping out over 50% of all entry-level white collar jobs across finance, technology, law, and consulting sectors, causing up to 20% unemployment in the next one to five years.
Companies are struggling to turn AI into profit
However, the companies rushing to weave AI tools into their routines reportedly haven’t had much success capitalizing on the technology. According to an MIT study, which surveyed 350 employees, interviewed 150 leaders, and analyzed 350 public deployments of AI, 95% of AI pilot programs fail to turn a profit or achieve performance goals, with only 5% of programs delivering rapid revenue acceleration or having a measurable impact on profits and losses.
Of course, an economic upheaval where entire job industries and roles become fully automated may still come, and it would be foolish to write off the tremendous potential AI holds this early. There’s no denying AI has had a clear and measurable impact on workplaces, but for it to evolve from just a tool that helps you complete tasks faster and more efficiently into something that can completely take over your job will be a significantly steeper hill to climb than early hype suggested.





