By Erin Dunlevy, VP and Consultant, True North: Center for Organizational Health
Recently, a former ed-tech client reached out with a dilemma.
For the last year, the company had been moving at a speed that would have been hard to imagine even a few years ago. AI was transforming products, accelerating development, reshaping user interfaces, and making it possible to build and iterate at a dizzying pace.
And then something happened that, in retrospect, probably should not have surprised anyone.
One of the company’s teams used AI to recreate an entire product with almost no human oversight. The new version was developed quickly, efficiently, and pushed toward the market.
The development team was thrilled.
The academic leadership team was horrified.
From the academic team’s perspective, the company had created new pedagogical materials without consulting the people responsible for the educational integrity of the product. There were questions about quality, ethics, values, and ultimately: Who gets to decide what good teaching looks like?
From the development team’s perspective, the reaction was baffling. They had found a way to save enormous amounts of time and effort. Why wouldn’t the organization want to take advantage of that? Why slow down a process that AI could make faster?
Both sides believed they were doing the right thing. And both sides weren’t necessarily wrong.
This is where the conversation about AI gets much more interesting. Because the problem wasn’t really that one team “liked AI” and the other didn’t. It wasn’t even necessarily about the technology itself. The problem was that the technology had moved faster than the organization’s ability to make values-driven decisions together.
The conflict underneath the conflict
When we look at organizational conflict, we often focus on the moment when everything goes sideways: the tense meeting, the angry email, the leadership retreat called to repair relationships.
But the conflict usually starts much earlier. In this case, each team was operating from a set of assumptions that made perfect sense within its own world. The development team assumed that innovation meant moving quickly. They saw a problem and found a more efficient way to solve it. Their mental model was: If we can build it better and faster, we should.
The academic team assumed that meaningful innovation required expertise, collaboration, and careful consideration of the human consequences. Their mental model was: If we’re changing something this important, the people who understand the educational impact need to be involved. Neither assumption is inherently unreasonable.
The trouble is that the organization had not created a shared agreement about which assumptions should govern when the two came into conflict.
And that is the part we think organizations are getting wrong about AI. We’re spending enormous amounts of energy asking, What can AI do? We should be spending at least as much time asking, How do we want to make decisions about what AI does?
Those are two very different questions. The faster the technology moves, the more important our human systems become.
The organizational habits that AI exposes
AI doesn’t create every organizational problem we’re seeing right now. Sometimes, it just shines a very bright light on the ones that were already there.