OpenAI Reports AI Boosts Daily Task Efficiency

The Impact of AI on Daily Work

OpenAI has revealed that its AI tools can save workers between 40 to 60 minutes per day across a variety of job roles. This claim is based on a survey conducted within real workplace environments, rather than controlled lab settings. The findings come at a time when many business leaders are still skeptical about the tangible benefits of AI technology.

The data highlights significant time savings in specific fields such as data science, engineering, communications, and accounting. These improvements focus on task speed rather than job displacement. OpenAI tracked how long routine tasks took with and without their software, noting that the most substantial gains were seen in writing, research checks, light coding, and document-related tasks.

The survey involved 9,000 employees from 100 companies, with responses collected from workers who had used the tools for three to four weeks. According to the results, 75% of participants reported that AI improved either their work speed or the quality of their output. Heavy users who combined advanced models with multiple tools experienced the most significant benefits, while casual users saw more modest improvements.

Time Savings Across Different Roles

Not all job roles experienced the same level of time savings. Workers in data science, engineering, communications, and accounting reported the largest reductions in task duration. In these roles, AI tools handled drafts, checks, summaries, and basic analysis in minutes instead of nearly an hour.

This finding comes amid lingering doubts about the value of AI investments. Earlier this year, researchers from the Massachusetts Institute of Technology found that most organizations did not see a return on their generative AI spending. Shortly after, teams from Harvard and Stanford noted that many professionals were producing what they called "workslop" — AI-generated work that appears polished but lacks depth.

These studies have fueled concerns about a potential tech bubble, as companies invest billions without clear paths to profitability. In response, AI firms like OpenAI have released their own reports to demonstrate the real-world impact of their technologies.

Enterprise Adoption and Usage Trends

OpenAI now claims over one million businesses are paying for its enterprise AI tools. Additionally, there are seven million paid users for ChatGPT in professional settings. The company states that business adoption is occurring at a pace comparable to, and sometimes faster than, consumer growth. OpenAI's Chief Operating Officer, Brad Lightcap, addressed the discrepancy between external research and internal data.

“There’s a lot of studies flying around saying this, that, and the other thing,” Brad said. “They never quite line up with what we see in practice.” He emphasized that enterprise use is expanding across teams, no longer confined to IT departments.

Usage patterns have also evolved. Employees who utilized advanced models and integrated multiple tools into their workflows saw the greatest productivity gains. Furthermore, the report indicates that people are using AI for tasks they previously avoided. For example, employees in engineering, IT, and research — who do not typically hold technical roles — showed a 36% increase in coding-related messages over the past six months.

New Capabilities in the Workplace

OpenAI’s Chief Economist, Ronnie Chatterji, linked this shift to new capabilities emerging within offices. “Three out of four people are now saying, ‘I can do things I couldn’t do before,’” Ronnie stated. He added that this aspect of the work transformation often goes unnoticed in broader AI discussions.

As AI continues to reshape the workplace, it is clear that the technology is delivering measurable benefits. However, the debate over its true impact remains ongoing, with companies and researchers offering conflicting perspectives. As more organizations adopt AI tools, the conversation will likely evolve, driven by real-world results and evolving usage patterns.