Follow ZDNET: Add us as a preferred source on Google.
ZDNET’s key takeaways
- A study from OpenAI aimed to assess the business impacts of AI.
- Productivity gains were found to vary between particular roles.
- Early studies like this should be taken with a grain of salt.
The use of AI is saving workers an average of 40 minutes to one hour of work per day, according to a new report from OpenAI.
The report, published Monday and titled “The state of enterprise AI,” focused on the use of ChatGPT Enterprise across a range of industries, and homed in on the roles and tasks benefiting most from the technology.
It arrives just a little over one week after ChatGPT’s third birthday, and during a time of growing competition among tech firms to corner the market for business customers, which OpenAI describes in the report as the key to unlocking the true economic value of AI.
“For much of the past three years, the visible impact of AI has been most apparent among consumers,” Ronnie Chatterji, the company’s chief economist, wrote in the report. “However, the history of general purpose technologies — from steam engines to semiconductors — shows that significant economic value is created after firms translate underlying capabilities into scaled use cases. Enterprise AI now appears to be entering this phase, as many of the world’s largest and most complex organizations are starting to use AI as core infrastructure.”
Also: OpenAI is secretly fast-tracking ‘Garlic’ to fix ChatGPT’s biggest flaws: What we know
(Disclosure: Ziff Davis, ZDNET’s parent company, filed an April 2025 lawsuit against OpenAI, alleging it infringed Ziff Davis copyrights in training and operating its AI systems.)
OpenAI, which kicked off the AI race with the release of ChatGPT, has also been facing mounting pressure to stay ahead. Anthropic’s valuation has skyrocketed thanks in large measure to its popularity among enterprise customers, and the recent advancements in Google’s own AI efforts have also reportedly caused OpenAI CEO Sam Altman to announce an internal “code red.”
Saving time at work
For the new study, OpenAI analyzed anonymous user data and conducted a survey of 9,000 workers across 100 organizations. ChatGPT Enterprise now serves more than seven million individual workers, and subscriptions have grown more than ninefold year-over-year, according to the company.
Three-quarters of survey respondents said their use of AI has boosted either the speed or quality of their work. On top of the 40-60 minutes of saved time reported by the average respondent, data science, software engineering, and communications workers said their efficiency gains have been even higher — between one hour and 80 minutes, on average.
Also: How AI-enabled autonomous business will change the way you work forever
The tasks most amenable to being sped up through the use of AI vary between roles, according to the report. For example, 85% of marketers reported faster campaign execution, while 73% of engineers said the technology was enabling quicker code delivery.
“These results indicate that productivity benefits are already materializing across core enterprise functions, not only in early-adopting technical roles,” Chatterji wrote, echoing a report from MIT published last week, which found that AI can already automate close to 12% of the total US workforce — not only across tech, but also in industries like HR and finance.
Yes, but…
Studying the use of AI in the workplace and translating that to concrete economic value remains a young and inexact science. Early reports like this one from OpenAI should therefore be taken with a grain of salt.
All AI developers have an obvious incentive to frame their technology and its economic impacts in a flattering light. But there are many factors pertinent to the broader conversation about the use of AI in the workplace that aren’t directly accounted for in OpenAI’s study, and that also merit closer study.
Also: Dreading AI job cuts? 5 ways to future-proof your career – before it’s too late
For one, it’s not clear that faster outputs enabled by AI are always translating to higher quality, or if this is just causing the tide of so-called workslop to rise. Another study published earlier this year by freelancer platform Upwork also found a correlation between the use of AI at work and burnout, suggesting that the technology could be taking a negative psychological toll on workers even as it boosts their productivity. And the technology could, as some industry leaders have readily admitted, displace a large number of human workers, an outcome that could produce wide-scale social, political, and economic repercussions.
For the time being, though, as they look towards securing ever-bigger slices of the enterprise customer pie, leading AI developers will probably continue to focus on the technology’s capacity to boost productivity, and to draw a direct line between that and broader economic prosperity.
A recent study from Anthropic, for example, analyzed 100,000 conversations with Claude and concluded that the chatbot was helping users complete tasks 80% faster, which, according to the company, meant that the use of current AI models could cause the US economy’s growth rate to double over the next decade.