OpenAI asserts that its artificial intelligence tools now save an average of 40 to 60 minutes per day across various job roles. This claim stems from a new survey conducted within real workplace environments, rather than controlled laboratory settings.
The findings emerge at a critical juncture, three years into the widespread adoption of AI, a period where many business leaders are still evaluating the tangible benefits of the technology.
Workers in fields such as data science, engineering, communications, and accounting reported the most significant time reductions. The survey's focus was on the acceleration of tasks, not on the potential impact on employment levels.
OpenAI monitored the duration of routine tasks with and without the integration of its software into daily workflows. The primary efficiencies, according to the company, were observed in writing, research verification, basic coding, and document management.
The survey encompassed 9,000 employees from 100 different companies. All participants had been using the AI tools in their professional capacity for three to four weeks prior to responding. A substantial 75% of respondents indicated that the AI tools improved either their work speed or the quality of their output.
Individuals who extensively utilized advanced AI models and combined multiple tools experienced the most substantial time savings. Conversely, casual users noted more modest improvements. OpenAI also analyzed the results based on job type and industry to identify specific areas where time efficiencies were most pronounced.
Task Time Reduction Across Key Roles
The extent of time savings varied across different professional roles. Jobs within data science, engineering, communications, and accounting demonstrated the most significant decreases in the time required to complete tasks.
In these specific roles, employees reported that the AI tools were capable of handling tasks such as drafting documents, conducting checks, summarizing information, and performing basic analysis in mere minutes, a process that previously could take nearly an hour.
This survey data is being released at a time when skepticism regarding the return on investment for AI remains prevalent. In August, researchers from the Massachusetts Institute of Technology suggested that most organizations had not seen any discernible return from their investments in generative AI. The following month, teams from Harvard and Stanford published findings indicating that many professionals were producing what they termed "workslop" – AI-generated content that appears polished but lacks substantive depth.
These studies have contributed to concerns about a potential new tech bubble, with companies investing billions without a clear path to profitability. In response, OpenAI and other AI companies have released their own reports to highlight the practical, day-to-day impact of their technologies on businesses.
Just last week, rival company Anthropic stated that its Claude model reduced task completion time by 80%, based on an analysis of 100,000 user conversations. It is important to note that none of these reports have undergone peer review. The ongoing debate about the true value and impact of AI is now being shaped by these public disclosures.
Enterprises Accelerate Paid Use of AI Tools
OpenAI now reports that over one million businesses are actively paying for its enterprise AI solutions. The number of paid seats for ChatGPT in corporate settings has reached seven million users. The company indicates that the pace of business adoption is now comparable to, and in some instances faster than, consumer adoption rates.
Chief Operating Officer Brad Lightcap addressed the discrepancy between external research findings and internal company data. "There’s a lot of studies flying around saying this, that and the other thing," Brad commented. "They never quite line up with what we see in practice." He further elaborated that enterprise usage is expanding across various teams within organizations, rather than being confined to IT departments.
Usage patterns have also evolved. Employees who leveraged advanced AI models and integrated multiple tools into their workflows reported the most significant productivity gains.
The report also highlights instances where individuals are using AI for tasks they previously did not engage with. Employees in engineering, IT, and research roles, who are not in traditionally technical positions, have shown a 36% increase in coding-related communications over the past six months.
Chief economist Ronnie Chatterji attributed this shift to the new capabilities available within workplaces. "Three out of four people are now saying, ‘I can do things I couldn’t do before,’" Ronnie stated. He emphasized that this aspect of the evolving work landscape is often overlooked in discussions about AI.