Here’s a question worth raising at your next leadership meeting: How many AI tools are being used inside your organization right now?
Not how many IT approved. Not how many the company pays for. How many are actually in use, today, across your teams?
If the room gets quiet, you’ve just found your AI strategy’s first blind spot.
Somewhere in your organization — probably in plain sight — Marketing is running copy through ChatGPT. Finance is building a Copilot workflow nobody signed off on. Customer Service found a browser extension that saves an hour a day. Someone in Operations built an agent over the weekend. And someone, somewhere, has definitely signed up for an AI tool nobody else knows exists.
Did any of this happen because IT rolled out a plan? Nope. It happened because the tools are fast, accessible, and often easier to start using than it is to ask for permission.
Everyone has AI now. That’s the good news … but also part of the problem.
The AI Sprawl Paradox
Definition: The faster people find useful ways to adopt AI, the faster adoption can outpace the organization’s ability to see it, govern it, and learn from it.
The same experimentation that signals an engaged, capable workforce can also create risks the organization cannot see.
At first glance, all that unapproved activity may sound like a governance failure. But look closer. It may be evidence that your people are doing exactly what you want them to do: finding faster, smarter ways to solve problems without waiting for someone to hand them a solution.
That’s not a culture problem. It’s a culture worth building on.
The issue isn’t that AI is spreading. It’s that the organization’s ability to see it, evaluate it, and learn from it isn’t keeping pace. People are adopting tools faster than anyone can track where, how, or why they’re being used. That’s the AI Sprawl-blem: a growing collection of powerful tools, full of potential, and also a recipe for chaos if nobody’s steering.
The sprawl itself isn’t the enemy. Unmanaged sprawl is. And hidden inside that sprawl is a real opportunity for organizations willing to lead through it.
The Upside Is Real
AI is already helping teams move faster. Work that once took a week can sometimes be finished in an afternoon. Employees can answer questions, test ideas, generate first drafts, summarize information, and build lightweight solutions without waiting on an already-crowded technical backlog.
The biggest shift is self-service. People can solve more of their own problems, on demand. One of business’s oldest bottlenecks begins to loosen, and employees gain more ownership over how work gets done.
That’s valuable. But it doesn’t automatically make every tool useful, secure, or worth keeping. The opportunity comes from separating productive experimentation from basic tool accumulation.
You Have to Experiment,
But You Also Have to Learn
In a landscape changing this quickly, nobody has the complete playbook. Not the vendors, not the analysts, and not your IT team. The only reliable way to discover what works is to try things.
And “what works” won’t be the same for every organization. Two companies in the same industry may reach very different conclusions. Different departments inside the same company may need different approaches. The right choice depends on the problem, the process, the people, and how confidently the tool can be used.
Right now, experimentation is part of the work. The better question is whether your organization is running experiments or simply accumulating tools.
An experiment has a purpose:
- A problem it’s meant to solve
- An owner who’s paying attention
- A defined group of users
- Clear boundaries around data and access
- Checkpoints for reviewing what happens
- A way to determine whether it’s working
Tool accumulation has none of that. Tools pile up. Teams solve the same problem three different ways. Nobody compares notes. Six months later, no one can say what created value, what introduced risk, or what’s quietly running in the background. The difference isn’t how much you try. It’s how intentionally you track and evaluate what you try. That takes planning.
There’s also another side to the adoption question. Some employees may hesitate to participate at all. If experimentation is happening everywhere except in certain teams or roles, leadership needs to understand why. The barrier may not be resistance to change — it may be a lack of time, training, trust, or clarity about what’s allowed. That’s not just a technology issue. It’s an adoption issue, and it deserves the same attention as the tools themselves.
The Hidden Cost of What You Can't See
For IT teams, self-service can provide welcome relief from a backlog that never seems to shrink. But that relief introduces a new challenge: visibility.
When anyone can spin up a tool, workflow, or agent, basic questions become surprisingly hard to answer:
- What’s currently in use?
- Who owns it?
- What problem does it solve?
- What data can it access?
- Which business process depends on it?
- What happens if it disappears?
These questions matter most when it’s time to consolidate, standardize, or retire something. Removing a tool without disrupting a business process is MUCH harder when no one knew the tool existed until something broke. The same is true when three teams discover they’ve solved the same problem three different ways.
There’s also a human cost. People get comfortable with the tools and processes they build, especially when those solutions make their jobs easier. They invest time, effort, and real ownership into making them work. Retiring one doesn’t always feel like a technical decision — it can feel like losing something they created. Those conversations take time, context, and care, and they often fall to managers and IT teams that are already stretched.
When Sprawl Becomes Risk
What’s being used? Who owns it? What data does it touch? What depends on it?
If leadership can’t answer those questions, visibility is already the first problem.
For growth-stage and mid-market organizations — especially those in regulated environments like financial services, insurance, healthcare, and nonprofits — invisible AI use is more than an inconvenience. It can become a compliance exposure, a data-governance gap, an operational dependency, and a strategic blind spot all at once.
When dozens of tools are operating across the organization with no central view of what they’re doing, you don’t have an AI strategy. You have AI sprawl. And unmanaged sprawl is where real risk lives.
Bring a Plan, Even If the Tools Are Already Here
Ideally, the plan arrives before the tools. If it didn’t, get one in place now.
The heart of that plan is visibility. You can’t steer what you can’t see, and you won’t see much if employees feel they need to hide what they’re doing. The goal isn’t to lock everything down. It’s to make experimentation visible, useful, and safe.
Know what’s in use. Maintain a simple, living picture of the tools being used, who’s using them, what problems they address, and what data they can access. This doesn’t need to become a massive administrative exercise — it just needs to be current enough to support good decisions.
Evaluate as you go. Give every meaningful experiment a goal, an owner, and a review point. Tools should earn their place or be retired intentionally, not stay embedded because everyone forgot they were there.
Avoid overly restrictive policies. Heavy restrictions rarely eliminate experimentation. More often, they push it out of sight. Clear boundaries are necessary — especially around sensitive data, security, and compliance — but guardrails should tell people where they can move safely, not make every useful idea feel impossible.
Encourage communication and sharing. Give employees a place to share what they’ve built and what they’ve learned. Good ideas can spread, duplicated efforts can surface earlier, and risks can be addressed before they become embedded in critical processes.
Explain the “why.” When a tool must be standardized, replaced, or retired, say why. Share what the organization learned, what it’s moving toward, and how the decision benefits the people affected. People accept change they understand. They resist change that just shows up.
With the right visibility and communication in place, leadership can make better decisions about what to encourage, what to standardize, and what to stop. But visibility does more than help manage the sprawl — it helps the organization learn from it.
The Opportunity Hiding in the Sprawl
Every AI tool someone adopts is a signal. It may mean:
- A process is too slow
- Information is difficult to find
- Repetitive work is consuming too much time
- An existing system isn’t meeting a need
- A team lacks access to the right expertise
- Employees see an opportunity leadership hasn’t noticed yet
Read that way, AI sprawl becomes a diagnostic tool for organizational friction. It shows where people are struggling, where existing processes are breaking down, and where employees believe work could be done differently. In some cases, those patterns reveal more than a survey or strategy session ever could.
That’s another reason visibility matters: you can only read the signals you can see. And if leadership pays attention, some of those experiments may point toward permanent operational improvements. The lasting value may not be the individual tool an employee chose. It may be the problem that choice brought into view.
As AI becomes more accessible, the advantage will increasingly come from how organizations apply it, evaluate the outputs, and incorporate human judgment into the result. The winners won’t be the ones with the most AI tools or the fewest. They’ll be the ones that experiment with intent, understand what’s happening across the business, recognize the signals, and lead people through change with a clear sense of why.
So, let’s return to the question we started with: How many AI tools are being used inside your organization right now?
If the honest answer is still “more than we know,” that’s your starting point. The goal isn’t to control AI or simply shrink the number of tools. It’s to make experimentation visible, learn what that experimentation is telling you, and turn grassroots problem-solving into deliberate organizational progress.
Let the rainbow of options inspire you. Just bring a plan.
Need Help Assessing Your Sprawl-blem?
If you’re working through any of this and want help building that plan or identifying opportunities to improve operations, contact me at 303.653.9717, email chris.wojciak@thecrmfirm.com, or visit thecrmfirm.com/contact and connect with The CRM Firm team.
