The Hidden Use of AI: Why Transparency Matters
As AI continues to gain momentum within companies, it’s important to distinguish between two main categories of AI usage: stationed, strategic projects, and the personal, often hidden, use of AI by individuals in their workflows.
One challenge in the latter category is that people are hiding the fact that they are using AI in their work. This fact has been well documented. Two primary drivers contribute to workers hiding their AI use. First, using AI carries a stigma that it isn’t their own work. This concern is valid, especially when AI is used without critical thinking, judgment, or fact-checking. But we’ve used aids, both human and technology, since the beginning of human history, so AI shouldn’t be any different. The other reason is that people are concerned that if productivity gains through the use of AI are made more visible, job cuts will be right around the corner.
Lean, or continuous improvement, went through a significant wave of this same concern. If we practice lean to make things more productive, won’t the company just lay off the excess? Some companies intended this very thing, and that made the whole problem worse. However, many companies had solid intentions but their stated intentions did not serve to appease the skeptics.
As a result, many companies decided to make a declared promise not to have any layoffs because of continuous improvement efforts. This was in vogue for quite a while and got lots of attention. Some reference Toyota as an example. Toyota’s no-layoff policy was more of a union concession after a layoff rather than the result of gains from the Toyota Production System. T lean efforts did make it far easier to comply with that promise by making the company stronger and more agile.
One of the reasons that the “no-layoff” practice faded is that the more accurate version of the promise was, “no layoffs because of continuous improvement until economic conditions require action.” Because of this, the promise lost its integrity. But companies, having made this the intent while introducing lean strategies, certainly made efforts to continue their practice.
Despite its decline, this practice was a useful mechanism to help contextualize the intent of lean efforts, which was to improve for more capability, more capacity, more engagement, and not just fewer employees. Perhaps this practice could be useful in the adoption of another new productivity enhancer: AI.
Sure, the commitment may not carry a tremendous amount of weight, but it can at least signal a company’s intention.
Why is this a worthy problem to solve?
First, with more conversation about the use of AI, it can remove the stigma of AI as “cheating.” AI is a tool to enhance both the quality and the quantity of output. Most technologies make the same promise but just lack the broad application of AI. By removing the stigma, adoption can accelerate.
Second, real progress is accelerated when best practices are shared. I learn a lot by sharing how I’m using the tools as well as by learning from others. Every tip, trick, and technique can add a few percentage points of progress.
Third, standard work is critical to consistent performance and problem solving, but it is only helpful if it is up-to-date and followed. If people modify the standard work “off the books” with use of AI, then your standard work promises to be out of date all the time, and fundamentally useless.
Making no-layoff commitments may not convince everyone that they are fundamentally safe. Leaders must be conscious of people’s fears and concerns. Remove the resistance to openly using AI to help accelerate your progress transparently and collaboratively.