46% Of Knowledge Workers Tell Their Bosses To Take A Hike When It Comes To GenAI

The firm Software AG just conducted a study of 6,000 employees that said 75% of knowledge workers are using GenAI and 46% would not stop using these tools even if the company forbade their use. This finding has profound implications for executives. It creates a massive security problem if rogue IT users share data with models and providers without review or approval, and more fundamentally it means that just about half your knowledge workers are not going to go back to old ways of working – no matter what you do!

GenAI Drivers Underneath the Mutiny

Why are knowledge workers so committed to this path? I think there are fundamentally three reasons:

1. There Are Massive Productivity Gains For Each Individual.

GenAI enables knowledge workers to complete tasks faster, automate repetitive work, generate ideas, and improve quality. When AI tools save them hours per day, banning them feels like a direct threat to their efficiency—and their ability to stay competitive in their roles.

2. Competitive Pressure and Career Survival

Workers see their peers (and competitors) using GenAI to get ahead. If they don’t, they risk falling behind in productivity, creativity, and innovation. Many believe that mastering GenAI is essential for their long-term job security and career growth.

3. Work Arounds Are Too Easy.

With consumer-grade AI tools available on personal devices, banning GenAI is almost impossible to enforce. Employees can access AI outside of corporate networks, making bans feel more like a bureaucratic hurdle than a real deterrent.

If You Can’t Beat’em, Join’em: Three Implications for Leaders

First, in a world where 75% of knowledge workers already use large language models, it is vital that organizations have a secure employee chatbot available, and that policies and access are easy and clear. If you make this process too sticky, people will work around it and create new potential security risks.

Second, top down policies and strategies should be augmented by incentive based controls. If you make tools, training, and other resources available to these employees who embrace these ways of working, they will have an incentive to share their activities. In the early days of personal computing, these types of incentive-based policies helped leaders keep up with adoption and use. The GenAI wave has been faster, and will be bigger, than personal computers.

Third, this type of committed adoption gives on the ground evidence that those doing the work see the benefits and value of these new tools. Given that we are less than two years into the introduction of GenAI broadly, I expect not only is GenAI here to stay, but it’s use is likely to accelerate. This gives firms a need to build capacity to productively introduce and use newer and newer models as the collective IQ increases across the technology stack. Remember, not only do these tools improve more rapidly all the time, it is also the first time in human history that we can implement new intelligence across and enterprise instantaneously without the long process of training and eduction.

By the end of 2025, every management team needs to understand that we are already in a world with regular workers and digital workers and the faster we realize that, the more we can develop the management and leadership pratices needed in this new world. It is only accelerating at an accelerating rate.

As Dorothy said to Toto in the Wizard of Oz, “We’re not in Kansas anymore.”