From 26% to 76% in One Year: Every Company Is Hiring a Chief AI Officer. IBM — the Same IBM from 1997 — Has None.

A few weeks ago, IBM published a study with data from 2,000 CEOs across 33 countries: the percentage of companies that now have a Chief AI Officer jumped from 26% to 76% in just twelve months. It’s probably the fastest growth any executive role has seen in recent corporate history.

Read that again: in one year, the number nearly tripled.

And here’s the twist that gives this column its point. The same IBM that published that study — the same IBM whose 1997 “e-business” campaign taught the corporate world to treat every new technology as a separate department — doesn’t have a Chief AI Officer. The person who fills that function, Joanne Wright, Senior Vice President of Transformation and Operations, put it plainly: “I don’t own AI.” Every leader of every area at IBM is responsible for how artificial intelligence is adopted within their own team. Wright’s role isn’t to control that adoption — it’s to remove friction from it.

It’s no coincidence that it’s IBM doing this differently. It’s the same company that in 1997 wrote the manual for how not to treat a general purpose technology, and thirty years later seems to have learned the lesson before the rest of the market.

This doesn’t mean hiring a Chief AI Officer is, in itself, a mistake. IBM’s own study found that companies with that role report a 5% higher AI return on investment than those without one. The problem isn’t the title. The problem is what that title is expected to solve alone, forever.

The same IBM study, and others published this year, point in the same direction: 85% of the CEOs surveyed say that all functional leaders will need to become technology experts within their own area. 79% of executives say they’re already decentralizing decision-making as AI expands within the company. And organizations that simultaneously redesigned five core areas — technology, finance, human resources, operations, and cross-team collaboration — are four times more likely to meet their business objectives than those that left everything in the hands of a single role.

The honest reading of these numbers isn’t that the Chief AI Officer role is going to disappear overnight. It will continue to exist in many companies, especially for regulatory reasons: the new European AI regulation, for example, requires someone to be formally responsible for the highest-risk systems. The honest reading is different: the role that survives isn’t the one that centralizes all AI decisions — it’s the one dedicated to governance and removing friction for the rest of the organization, while each area — marketing, finance, operations — becomes the owner of its own adoption.

Schneider Electric, another company cited in the IBM study, gave this model a name: they call it “hub and spoke” — a small center that defines common standards and a set of areas that execute with autonomy within those standards. It’s neither the centralized office everyone is building today, nor the chaos of every area inventing its own rules. It’s something in between, designed on purpose.

If your company has already named someone in charge of AI, the question worth asking isn’t whether that decision was wrong. It’s whether that person’s mandate has an explicit expiration date: whether their job is to build the standards and then hand responsibility to each area, or whether — without anyone having consciously decided this — it’s becoming the only place in the company where AI actually lives.

Was the AI role in your company designed to become unnecessary, or to stay forever?