The Vendor Question Nobody Asked During the Rush

The Deliberate AI Leader — A Series for Executives Who Want to Get This Right – Part 18

The Question That Gets Skipped

When a team adopts a new AI tool, the questions in the room are usually about capability. Can it do the task. Is it fast enough. Does the team like using it. Almost nobody asks what happens if this vendor changes their pricing, gets acquired, or simply stops being the best option in eighteen months. That question doesn’t feel urgent during a demo. It feels very urgent during a contract renewal, once three other systems have been built on top of the decision.

The Numbers Are Worse Than Most Leaders Assume

Recent research from IBM found that a large majority of executives, seven in ten, said switching their primary AI vendor or model would be genuinely difficult. An even larger share, more than nine in ten, admitted they don’t have complete visibility into their organization’s AI dependencies in the first place. Put those together and you get a common pattern: companies that couldn’t tell you today everything their operations quietly rely on, and would struggle to untangle it if they had to.

This isn’t a procurement footnote. It’s becoming an operational resilience issue, the same category as relying on a single supplier for a critical part.

How Lock-In Happens Without Anyone Deciding It Should

Nobody sits down and chooses vendor dependency on purpose. It accumulates through a series of individually reasonable decisions. A team picks a chatbot because it’s already approved. An automation gets built to route data through it. A second workflow gets layered on top because the first one worked. Eighteen months later, three departments have processes that assume that specific tool exists, in that specific form, indefinitely.

Our earlier post on the three tiers of AI, Chatbots, AI Agents, and Automation: What You’re Actually Buying, covers how automation platforms connect tools into an operations stack. That same connective tissue is exactly what makes unwinding a single vendor so painful later. The more layers built on top, the more expensive the dependency becomes to see, let alone remove.

It also tends to hide inside the workarounds people build to cover for a system’s gaps, which is the pattern we described in Human Middleware Was Never a Scalable Operating Model. A person quietly patching a tool’s limitations is also quietly making that tool harder to replace, because now the fix lives in someone’s head instead of the system.

A Preference Is Not the Same Thing as a Dependency

Using a tool because it’s genuinely the best option is fine. That’s just a good decision. The problem is not having a way to tell the difference between that and being stuck. Three questions separate the two:

• Could we export our data and history out of this system today, in a usable format, without vendor cooperation?
• If this tool disappeared or doubled its price next quarter, do we know which specific workflows break, and who would notice first?
• Is there a documented fallback, even a manual one, or does the plan currently amount to hoping this never happens?

If the honest answer to any of those is no, that’s not a crisis. It’s information. It tells you where to spend the next hour of governance work.

Why This Matters for the Bake-Off

In What Survives the Bake-Off, we talked about the sorting happening across every AI initiative right now. Vendor entanglement is one of the biggest reasons that sorting stalls. A tool that should have been retired or replaced stays in place because unwinding it feels harder than it actually is, and nobody has taken the hour it would take to find out.

Organizations that adapt fastest treat vendor relationships as replaceable by design, not as permanent fixtures. That’s a version of the same adaptability gap we described in The Adaptability Gap: What AI Is Really Exposing Inside Your Organization, where the constraint was never really the technology. It was how tightly the organization had welded itself to one way of doing things.

Where to Start

Pick your three most business-critical AI tools. For each one, answer the three questions above honestly, in writing, before the next renewal date rather than during it. That single exercise will tell you more about your actual AI risk than any vendor’s roadmap presentation.

If you want help mapping where your operations are more dependent than you’d like, book a Strategy Call. We’ll help you find the entanglements before a vendor forces the conversation.

About WHIM Innovation

WHIM Innovation helps organizations harness the practical power of AI, automation, and custom software to work smarter and scale faster. We combine deep technical expertise with real-world business insight to build tools that simplify operations, enhance decision-making, and unlock new capacity across teams. From AI strategy and workflow design to custom monday.com apps and fully integrated solutions, we partner closely with clients to create systems that are efficient, intuitive, and built for long-term success.