Council Post: Why 'We Own It' Is The C-Suite's Favorite AI Lie
tags:Florian Douetteau is co-founder and CEO of Dataiku, a Platform for AI Success.
getty
Here today, gone tomorrow. Who would have thought the song by the Ramones in the 1970s would be relevant when it came to talking about AI? And yet, it perfectly describes what happened last month, when a government order took the most capable AI model on the market offline within hours. A decision was made, and the model was switched off—for everyone, everywhere, instantly.
Across Europe and Asia, this conversation has been happening in parliaments, boardrooms and think pieces for years. In the US, we haven't named it yet, but the problem is identical. Call it AI Sovereignty or AI Independence, but the question remains: when someone else can switch off the intelligence your business runs on, do you really own it?
If that happens, no contract you signed with your vendor stops it. This isn’t a technology risk, but a control problem. And most enterprises are already living inside a version of it—they just haven't had their moment yet.
Paying Rent On The Core Of Your Own Operation
The one thing most CEOs won't say out loud is that they don't own their AI. They rent it. Someone else built it, someone else can modify it and what the shutdown of Anthropic’s Fable 5 proved is that someone else, the government in that case, can take it away.
That risk didn't start with a government order; it was already there. A vendor raises prices, deprecates a feature you built on, shifts the roadmap and now you're locked in because leaving costs more than staying.
Our recent “Global AI Confessions Report: CEO Edition, 2026,” produced from a survey of 900 CEOs conducted in partnership with Harris Poll, makes the stakes visible. The study found that 70% of CEOs claim they are the primary driver of AI strategy within their organizations, yet only 6% are involved in nearly all decisions that determine whether that strategy succeeds or fails. That is not a strategy—but the appearance of one. And it's happening in an environment where, according to a 2026 KPMG US CEO Outlook Pulse Survey, 48% of CEOs don't feel there is sufficient market certainty to make significant investment decisions. Yet those CEOs don't view AI investments as discretionary but as required.
That is the definition of dependency built on a foundation nobody controls.
Personal Stakes
Our study also found that 80% of CEOs say their roles will be at risk if their companies fail to deliver measurable AI outcomes by year's end. They believe a fellow CEO will be ousted in 2026 due to a failed AI strategy or an AI-driven crisis. Compounding that pressure, approximately 60% of CEOs believe their boards are rushing AI transformation, according to the BCG CEOs and Boards Survey, pushing for speed on top of a foundation nobody fully controls.
What Ownership Actually Requires
When most CEOs say they own their AI, they mean the infrastructure: a sovereign cloud, a local model, the compliance boxes checked. That is the easiest level to point to, and the smallest part of what ownership means. Owning the building your AI runs in tells you little about whether your people understand the model inside it, whether they can change it or whether the intelligence doing your hardest work is still rented from a provider who can reprice it tomorrow or be ordered to cut you off. Infrastructure is the floor of ownership, necessary and far from the whole of it.
Real ownership operates at three levels, and infrastructure is only the first.
The second is capability: who builds the system, who understands it, who can modify and govern it. If your team cannot account for how a model reached a decision, you are a tenant with operator access, regardless of what the contract says.
The third is economic: who captures the value that AI produces, and whether you retain the freedom to change course as the market moves. The numbers say the mood has already turned. Two years ago, the worry was falling behind; today, CEOs say they fear over-investing in the wrong vendor more than moving too slowly, a clean reversal. Revenue growth has become the leading measure of AI success. None of that value is yours to keep if the pricing is opaque, the consumption unpredictable and leaving is engineered to cost more than staying.
Each level depends on the one below. The day you lose optionality is the day someone else starts making your choices for you.
Own What You Cannot Afford To Lose
Enterprises most at risk are not the ones that moved too slowly, but the ones that gave away the keys to their company and called it a strategy. The data suggests CEOs know it. When asked what matters most for AI success, they now rank governance ahead of both people and technology. Not capability or speed, but control. That’s because you only understand what ownership costs when you don’t have it.
The fix is not a different vendor, but a new perspective. Enterprises should use frontier models where they need the edge. Their own layer where their decisions actually live, the logic, the institutional knowledge, the processes your business depends on—that should be built on foundations you control, on models you can inspect, modify and keep. Models commoditize, and what doesn't is the judgment of your people, encoded into systems that answer to you.
Anthropic’s Fable 5 shutdown was a reminder, not an anomaly. Sure, right now it's being discussed as a geopolitical issue, but it goes deeper than that. AI Sovereignty is about ownership. And that question of who actually owns the AI will come up again and again for different models, vendors and reasons. The enterprises that will weather the storm won't be the ones with the best AI, but those who actually own it.
The only AI worth building: one that answers to you.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?