Spending up, confidence down

Two facts that don't seem to belong together. AI budgets are climbing fast — many companies are increasing AI spend by double digits, some doubling it year over year. And at the same time, 51% of CIOs say AI adoption is already moving too fast for them to manage, with a widening gulf between what their companies want from AI and the governance and skills they have to actually run it.

Read that again, because it's odd. The people approving the spending are telling you the thing they're funding is outrunning them. That's not a healthy way to deploy serious money, and the obvious conclusion — "everyone should slow down" — turns out to be the wrong one. The interesting part is what separates the teams who feel out of control from the ones who don't. Let me get into it.

"Too fast" isn't really about speed

When CIOs say AI is moving too fast, it sounds like a complaint about pace. It isn't. The same reports show the problem isn't the speed itself — it's speed without the things that make speed safe: governance, clear alignment to the business, people who know how to run it. Fewer than half say their AI strategy is even aligned with their wider business plan. The discomfort isn't "we're going quickly." It's "we're going quickly with no hands on the wheel."

That reframes the whole thing. A car at speed isn't dangerous because it's fast; it's dangerous if it has no brakes and no steering. "Too fast to manage" is a confession that the management — the brakes and steering — was never built. The spending bought the engine and skipped the controls.

Governance is the accelerator, not the brake

Here's the counterintuitive finding, and it's the one worth keeping. The teams that invested in governance aren't the slow, cautious ones holding everyone back. They're the ones moving fastest, because they have the confidence to scale. When you can see what your AI is doing, trust that it's aligned, and catch it when it drifts, you can step on the gas — because you'll know if something goes wrong. Without that, every step forward is a step into the dark, and sensible people hesitate in the dark.

So governance isn't the thing that slows AI down. It's the thing that lets you speed up without flying blind. The companies feeling out of control didn't move too fast. They moved fast and skipped the part that would have let them keep moving fast safely — and now they're scared of their own momentum.

What to do with this

Whether you're running AI at a company or shipping it solo, the takeaway is the same, and it isn't "go slower":

• Match your pace to your control, not to the news. The right speed isn't the industry's speed or your competitor's. It's the speed at which you can still see what your AI is doing and step in.
• Build the guardrails as you scale, not after. Logging, monitoring, a clear way to catch and correct bad output — these aren't bureaucracy. They're what let you go faster without fear.
• Spend on control, not just capability. If your AI budget is all model and no governance, you're buying the engine and skipping the brakes. That's exactly the combination the nervous CIOs are describing.

The goal isn't to be cautious. It's to earn the right to be fast.

The bottom line

The spend-up, confidence-down paradox isn't a sign that AI is moving too fast in some abstract sense. It's a sign that a lot of money went into capability and not enough into control, and now the people in charge can feel the gap.

The fix for "too fast to manage" isn't slowing down — it's building the management, because the teams with real governance are the ones moving fastest. Don't pace yourself to the herd. Pace yourself to what you can actually see and steer, and invest in the controls that let you open the throttle without driving blind.