A new OpenAI study has analyzed how over 1,500 organizations use ChatGPT in their day-to-day operations. The findings highlight an intriguing question that many executives are still putting off.
TL;DR
- A new OpenAI study (arXiv, August 2026) analyzes data from over 1,500 organizations and 17 million messages. The use of ChatGPT Enterprise increased sevenfold between June 2025 and March 2026.
- AI usage spans all levels of the organizational hierarchy. The highest intensity is among entry-level employees: They send about eight to nine more messages per week than the average active user in the same company. Managers and executives use AI less frequently.
- In Germany, this trend is even more pronounced: According to Ipsos, only about one-third of executives use AI regularly.
- This article provides context for AI usage in companies, explains why leadership becomes the key driver in this context, and outlines concrete steps you can take to stay connected with your own team.
For months, the central question in companies was: Is AI actually being adopted across all teams? That question has now been answered—at least according to one study. It is being adopted—faster than most people think. The new question is: Do leaders actually know what their people use AI for on a daily basis and how well they’re doing it?
Human-AI Leadership describes leadership that consciously integrates AI into decision-making, collaboration, and skill-building without relinquishing human judgment. Leaders actively shape how AI is used within the team, define its limits, and ensure that people retain their own capabilities in the process.
What the OpenAI data from the 2026 study reveals
The study “How Organizations Use AI: Evidence from ChatGPT” by Aaron Chatterji, David Holtz, and colleagues is so insightful because it isn’t based on self-reported data. It links real-world usage data from ChatGPT Enterprise to roles, task types, and business metrics, collected between January 2024 and March 2026 (arXiv, 2026).
The first indicator is sheer speed. The aggregate output volume of Enterprise customers increased sevenfold between June 2025 and March 2026. About half of this growth occurred within companies that were already using AI.
In other words: Once AI gains a foothold, it continues to spread—without any central rollout initiative.
AI usage cuts across the organization
A common assumption is that AI is something for the IT department. The data and common sense tell a different story. Its use spans many functions and multiple hierarchical levels, rather than being concentrated among the youngest employees or top executives. Among weekly active users, managers and directors actually make up the largest group at around 24 percent, while executives at the board level account for about 10 percent. So leadership is definitely involved.
The scope of AI usage is also broad in terms of content. More than half of active users perform documentation or writing tasks at least once a week, and just under half do technical digital work. Added to this are research, market analysis, planning, legal matters, and evaluations. AI has arrived at the very core of knowledge work.
How intensive is AI usage in companies?
The interesting finding lies in the depth of usage. Entry-level employees and trainees send about eight to nine more messages per week than the average active user in the same company. Managers, directors, and executives fall below this average.
Those who are new to the job make AI a natural part of their day-to-day work. Those in leadership roles use it less frequently.
Let’s put it more directly here: Your teams are often already further along in practical application than you are. Leadership is certainly involved, but AI is becoming part of the daily routine primarily among younger employees. And it is precisely this routine where expertise is developed.
The blind spot in leadership
In Germany, this disparity is even more pronounced. A survey by Ipsos commissioned by the Liz Mohn Foundation shows that only about one-third of executives use AI regularly. When it comes to challenging leadership tasks, the figure drops even further: 23 percent use AI for conflict resolution, and 19 percent for personnel decisions (Ipsos, 2025).
This creates a dangerous gap. While the workforce is integrating AI into their daily work, for many executives, what is actually happening remains abstract.
Which decisions are already being prepared by AI?
Where are decisions being made blindly, without scrutiny?
Who is doing it right, and who is just going through the motions?
These questions cannot be answered if one has had little hands-on experience with AI.
The consequence is evident in another statistic. In a Roland Berger survey, 59 percent of respondents believe their own leadership team is not sufficiently prepared for the AI transformation (Roland Berger, 2026). So leadership certainly recognizes its own gap—it just isn’t closing it.
You may know how this feels from your own meetings. A presentation seems surprisingly good; an analysis is completed astonishingly quickly. People nod approvingly and rarely ask how the result was arrived at. As long as the quality is there, everything seems fine. But without asking that question, it remains unclear where AI is being managed well and where it’s quietly producing errors that no one checks anymore. Leadership that doesn’t know how to use the tool itself loses the ability to make precisely this distinction.
Why leadership decides the course of AI transformation
There is no shortage of money or ambition. The BCG AI Radar 2026 reports that 72 percent of CEOs surveyed today see themselves as the primary decision-makers for AI—twice as many as in the previous year (BCG, 2026). AI has become a top priority. Nevertheless, implementation is stalling.
Roland Berger sums up this disconnect: 62 percent of executives expect AI to bring about major or radical changes to their business models, but only 38 percent have even begun the corresponding transformation. In Germany, according to an AWS study, 63 percent of companies use AI, but only 15 percent use it in a truly transformative way (AWS/Strand Partners, 2026).
Why is that? The Deloitte Global Human Capital Trends Report 2026 offers a clue. 85 percent of executives consider the ability to adapt quickly to be critical, but only 7 percent see themselves as leaders in supporting their workforce through this continuous evolution. Only 27 percent believe their organization manages change well (Deloitte, 2026).
Taken together, these figures paint a clear picture. The technology is here, the budget is in place, and the workforce has been using it for some time. What’s missing is leadership that actively shapes how AI is used, rather than letting it happen on its own. That’s precisely why leadership—not the tool—determines the success of AI transformation.
Where does your company really stand? The AI Readiness Check from triangility reveals in under 20 minutes how ready your organization is for AI, across four dimensions—from technology to culture—and provides concrete recommendations for action. Free of charge, available in German and English.
What leaders can do right now when it comes to AI
You can catch up—without having to become an AI expert overnight. It comes down to four specific steps.
1. Try it out yourself
Leading AI adoption requires using it yourself. This is essential for developing a sense of what the tool can do—and where it can be misleading. Take a recurring task of your own—such as preparing a decision-making document—and consciously work on it using AI for a week. No demo can replace this hands-on experience. If you want to approach getting started in a structured way, check out the seminar AI for Executives.
2. Make AI usage transparent
You can’t manage what happens behind closed doors. Make the actual use of AI within the team a topic for discussion. What is it being used for? Where does it help? Where do risks arise from unchecked automation? To this end, triangility works with the Gap Canvas—a tool for assessing current status—and a phased model consisting of diagnosis, analysis, and implementation. The core principle is simple: first understand what’s actually happening, then steer the process in a targeted manner.
3. Protect expertise, don’t erode it
The more cognitive work shifts to AI, the greater the risk that people will lose their ability to make judgments. We’ve described this effect elsewhere as Hollowing of Work: Jobs remain, but they lose their cognitive core. Leadership must actively ensure that the final judgment remains with humans and that employees continue to hone their skills rather than gradually relinquishing them.
4. Align Structure and Culture
Individual AI capabilities fizzle out if structures and culture don’t follow suit. Clear guidelines for working with AI, space for experimentation, and a mindset that allows for mistakes during the learning process determine whether numerous individual applications will lead to genuine transformation. The 17 New Leadership Principles describe how leadership remains sustainable in precisely this kind of uncertainty. The AI Transformation Program brings these building blocks together for organization-wide implementation.
We explore in depth how these four movements come together to form a human-centered leadership approach in our article AI and Leadership as well as in the eight typical barriers to AI transformation.
Get regular access to exclusive insights from thought leaders and practical tools on topics such as digital evolution, cultural transformation, future-readiness, resilience, mindfulness, and the design of new work environments.
To a lot of people, this study reads like a warning. I see it as an invitation. If teams have long since started using AI as a matter of course, we as leaders have the best opportunity in years to reconnect with the real work. That is, as long as we get hands-on with the technology ourselves and stay curious, rather than just trying to control things.
Conclusion: Adoption is here; now it’s up to leadership
The OpenAI study confirms what many leaders suspect but few are willing to say. AI has become part of everyday work life, driven by employees – often more intensely than by leadership itself.
At triangility, we’re convinced that the decisive factor is human. Technology spreads on its own. Whether it creates value depends on how leaders interact with the people around them: whether they themselves remain curious, make real-world use visible, and ensure that people grow – rather than being sidelined – when using AI. This is precisely where our understanding of Human-AI Leadership comes in.
View the gap in usage within your teams as a starting point. Leadership that addresses this transforms a vague everyday habit into a genuine, human advantage.
Frequently asked questions about AI transformation and leadership
What does the OpenAI study “How Organizations Use AI” reveal?
The study shows how companies actually use artificial intelligence. It analyzes real-world usage data from ChatGPT Enterprise across more than 1,500 organizations and 17 million messages, collected between January 2024 and March 2026. Key findings: Usage increased sevenfold between June 2025 and March 2026; it spans all hierarchical levels, and usage is highest among entry-level employees.
Do executives use AI less than their employees?
Executives are widely represented in terms of participation. However, they lag behind in intensity: According to OpenAI, entry-level employees send about eight to nine more messages per week than the average, while executives send fewer. In Germany, according to Ipsos, only about one-third of executives use AI regularly, and significantly less often for complex tasks such as personnel decisions.
Why is AI transformation a leadership task?
Because technology and budget are rarely the bottlenecks. Studies by Deloitte and Roland Berger show that implementation is hindered by a lack of adaptability, unclear governance, and insufficient preparation among leadership teams. The key requirements are therefore organizational, not technical. AI spreads on its own, but whether it creates value depends on how leadership shapes its use, expertise, and culture.
What does Human-AI Leadership mean?
Human-AI Leadership describes leadership that consciously integrates AI into decision-making, collaboration, and skill-building without relinquishing human judgment. Leaders determine what AI is used for, where its limits lie, and how people retain their capabilities. The approach combines practical AI use with clear responsibility for people and the organization.
How can leaders determine where their company stands with regard to AI?
A structured assessment quickly provides clarity. In under 20 minutes, triangility’s AI Readiness Check evaluates an organization’s readiness for AI across four dimensions – from technology to processes to culture – and delivers concrete recommendations for action along with benchmarks. It’s free and serves as the first step toward a targeted AI transformation.
What tasks do employees most frequently perform using AI?
According to OpenAI, more than half of active users of generative AI technologies such as ChatGPT perform documentation and writing tasks at least once a week, while just under half perform technical digital work. Other common tasks include research, market analysis, document creation, planning, and data and financial analysis. These AI applications are firmly embedded at the core of knowledge work, across all functions.
What is the greatest risk if leadership fails to manage AI usage?
The greatest risk is the gradual loss of judgment. As more and more cognitive work shifts to AI and no one is in control, people retain their positions but lose the cognitive core of their work. We call this effect the “hollowing of work.” Its effects are immediately visible in the workplace and, in the medium term, in the labor market as roles lose their value. Leadership must ensure that the final judgment remains with humans and that expertise continues to be developed.
What are the most common reasons AI projects fail in companies?
The implementation of AI technologies in the workplace often fails due to people, structures, and culture: unclear responsibilities, a lack of broad-based expertise, and leadership that does not actively shape the change. That is why sustainable AI transformation starts with leadership and organization – not just the tool itself.
Sources
- Chatterji, Holtz, Rakholia, Tambe, Weeratunga: How Organizations Use AI: Evidence from ChatGPT ()
- Ipsos im Auftrag der Liz Mohn Stiftung: Führung in Zeiten von KI ()
- Roland Berger: The AI-First Organization ()
- Deloitte: Global Human Capital Trends 2026 ()
- BCG: AI Radar 2026 ()
- AWS / Strand Partners: Unlocking AI Potential in Germany 2026 ()
Develop AI leadership skills: The only learning journey for leaders that combines artificial intelligence, new leadership, and AI transformation.
Join our New Leadership Community:
We send you our monthly newsletter on leadership, culture, organization and technology. With exciting, curated inspiration for the new world of work.
Contact
Contact Verena for personalized information on how to become more mindful as a leader.
You are currently viewing a placeholder content from Zoho Forms. To access the actual content, click the button below. Please note that doing so will share data with third-party providers.
More Information
