In the three years since AI went mainstream, I’ve watched transformation after transformation stall in the exact same spot. Not on budget. Not on technology. On people who quietly slow-walk the change they were hired to lead.
The easy explanation is fear. People are afraid of being replaced, so they drag their feet.
That explanation is comforting because it turns resistance into a character flaw, something to fix with better communication and one more town hall. It is also wrong, or at least dangerously incomplete.
The real explanation is structural. Once you see it, you cannot unsee it in your own organization.
I wrote about it in Enterprise Coherence and called it the tool-identity trap. AI didn’t create it. AI just made ignoring it expensive.
People Identify With Tools, Not the Work
Ask an accounts payable specialist what they do and most won’t describe the purpose of the job. They’ll name the system. “I’m an SAP expert.” Ask a data engineer the same question and you get “I’m a Databricks specialist” or “I run our Snowflake environment.”
Almost nobody says, “I make sure suppliers get paid accurately and on time so the business never loses a vendor relationship,” or “I make sure the business can trust the numbers it decides on.” They name the platform.
This isn’t a lack of self-awareness. It’s rational behavior in response to how the labor market actually works.
Recruiters search resumes for tool keywords, not process fluency. Certifications are built around platforms, Snowflake, Databricks, Salesforce, SAP, not around the flow of work those platforms support. Promotions go to the person who mastered the system, not the person who understands why the system exists.
Every incentive in a career points toward identifying with the tool.
So when a leader proposes changing the macro process behind a job, the leader hears “efficiency gain.” The employee hears something else entirely: the tool I built my identity and market value around for the last five years is about to stop mattering.
That isn’t a misunderstanding of the business case. That’s a full understanding of it.
Why AI Makes This Worse, Not Better
Every prior wave of technology change mostly left the tool intact and changed how you used it.
AI is different.
It’s starting to absorb the work the tool was for, not just the screen you clicked through to do it.
When an agent starts doing what the SAP expert used to do inside SAP, or what the Snowflake specialist used to do writing queries, there’s no version of the reassurance conversation that actually reassures anyone.
The threat isn’t abstract. Your most capable people can see it faster than any slide can explain it away.
That’s why resistance shows up hardest in your best, most tenured people. They aren’t confused about what AI does. They built an entire career on total command of one platform, and total command of a platform is exactly what AI starts to make optional.
The Fix Nobody Wants to Commit To
You cannot talk your way out of a structural incentive. No amount of change management messaging will convince someone their expertise is safe when their resume, their certifications, and their next promotion were all built on the thing you’re replacing. The fix has to be structural too.
Three moves work. None of them are hard to describe. All of them are hard to commit to.
Change what gets promoted.
Stop rewarding platform certifications and start rewarding people who can describe how value actually moves end to end, whatever that flow is in your business. When “understands Order-to-Cash from quote to collection” outranks “advanced Salesforce certification” on a promotion packet, behavior follows fast.
Create a role that owns the process, not the system.
Call it a Process Architect. Give someone explicit, permanent accountability for the integrity of a macro process, independent of which tools sit underneath it. This role should outlive every platform migration you’ll ever run.
Rotate your experts across the flow, not just up the ladder.
A finance analyst who understands the order side of order-to-cash stops fearing the next system change and starts advocating for it, because they can finally see the whole picture instead of just their piece of it.
The CIO’s Real Job Here
None of this is about being kind to people who are anxious. It’s about recognizing that your organization built a system, over decades, that rewards tool mastery over process understanding, and that system is now working directly against the transformation you’re trying to lead.
The organizations that get through this well aren’t the ones with the best AI. They’re the ones that changed what they promote before they changed what they automated.
Your best people are not resisting AI. They are protecting the only thing the labor market ever told them mattered.
Give them something better to protect, and the resistance stops being rational.
