Share with your CIO
Only 14% of enterprises deploying AI report having a clear strategy with defined goals and outcomes, according to a survey of more than 500 tech leaders by Altimetrik and HFS Research. The remaining 71% are operating with incomplete or developing plans, yet they’re deploying anyway. CIOs and CTOs hold accountability most often, and they feel pressure to ship before governance, training, or ownership structures exist. Nearly 80% of respondents say employees get fewer than 10 hours of AI training per year, and 43% report self-doubt when asked to use the tools.
What this means for your business
The gap between deployment pressure and strategic readiness isn’t evenly distributed. Enterprises that are still in early experimentation mode actually have an advantage here, because they haven’t yet locked in accountability gaps as permanent operating procedures. The organizations most exposed are the ones that have already shipped pilots without ownership structures, because probabilistic AI systems, unlike traditional rule-based software that only does exactly what it was programmed to do, fail in ways that don’t map cleanly to existing IT escalation paths.
Mark Baker’s line, “cost cutting is an outcome, not a strategy,” is sharp and worth defending. The recurring failure mode in enterprise technology adoption looks like this: a board-level imperative arrives, a mandate trickles down, and the team reverse-engineers a problem to fit the solution they’ve been told to buy. The survey data backs it up. Cost reduction as a primary AI driver is a symptom of that reversal. When the problem definition comes after the vendor selection, the ROI calculation is almost always post-hoc rationalization dressed as analysis. The companies KPMG identifies as seeing higher returns from mature AI deployments didn’t get there by chasing the cheapest outcome first; they started with a specific problem and built accountability around it.
The workforce finding deserves more weight than it’s getting. More than half of respondents expect roles to shift or shrink in the next few years, but most expect it to happen through attrition rather than deliberate planning. That’s not a neutral position; it’s a choice to let change happen to the organization rather than through it. The CIOs who get ahead of this won’t be the ones who run the most pilots. They’ll be the ones who can tell their board exactly which roles are changing, why, and what the transition looks like. I’d revise this view if we saw evidence that attrition-led AI workforce transitions were producing measurably better outcomes than planned ones, but the KPMG data pointing to investment in people as the differentiator for mature deployments runs the opposite direction.
Based on reporting from Most enterprises lack a clear AI strategy but push to forge ahead, originally published 2026-04-08 03:00:00.
