IBM: Why Most Enterprises Fail to Unlock AI Productivity

IBM has put its name behind research concluding that most enterprises have not converted their spending on artificial intelligence into measurable productivity gains, according to a report indexed by Google News on Sept. 24, 2026, under the headline "Why most enterprises fail to unlock AI productivity - IBM." The full methodology and the specific breakdown of causes sit inside IBM's original report, which readers can reach through that link; the aggregated headline itself is the confirmed fact as of publication.

What is the gap IBM is describing?

The headline points to a mismatch familiar to anyone tracking enterprise technology cycles: companies are buying AI tools faster than they are seeing productivity show up in output. IBM's framing — "most enterprises fail to unlock AI productivity" — puts a name on a pattern that has been building for several budget cycles, in which adoption of a tool is treated as a proxy for the return on that tool, even when the two numbers move independently.

"Why most enterprises fail to unlock AI productivity" — headline of the IBM-linked report, via Google News

That single line is doing a lot of work. It does not, in the material available at publication, specify which industries, company sizes, or use cases IBM studied, so readers who need the underlying data set — sample size, survey methodology, or country breakdown — should go to IBM's own report rather than to secondary coverage.

How much are companies actually spending on this?

The IBM finding lands as enterprise spending on AI productivity software keeps climbing. A market analysis reported by AI Productivity Hub projects the AI productivity tools market will reach $69.22 billion by 2035. That trajectory is the backdrop against which IBM's "failure to unlock" framing matters: if spending keeps rising while reported productivity gains stall, the gap between investment and return widens rather than closes on its own. Vendors selling into that market have every incentive to keep signing contracts regardless of whether buyers can prove a productivity lift, which is part of why independent research on the adoption gap — the kind IBM is reportedly publishing — carries weight for corporate technology and finance leaders deciding where the next budget cycle goes.

Why would spending more money produce less measurable gain?

Separate from the IBM report, other organizations tracking enterprise AI rollout have focused on the human side of adoption rather than the technology itself. AI Productivity Hub has reported on research highlighting the emotional dynamics of AI in the workplace, an angle that speaks to a related but distinct question: even when a tool works as designed, employee trust, anxiety about job displacement, and uneven training can determine whether a workforce actually uses it in ways that show up in output metrics. IBM's report and that emotional-dynamics research were not published as a joint study, and the aggregated material available does not show IBM crediting workforce psychology as a specific cause of the productivity gap it describes. The two lines of research are presented here as parallel context, not as the same finding.

Where can readers get IBM's full findings?

The most direct path is the original story surfaced through Google News, linked above, which credits IBM as the source. Readers looking for primary-document detail — IBM's survey questions, respondent counts, or industry segmentation — should look for IBM's own published report or IBM Institute for Business Value materials rather than relying on the headline alone, since aggregation services typically carry only a title and short summary rather than the full text.

What should enterprise leaders watch for next?

The practical follow-up questions for any company reading the IBM headline are straightforward: whether the productivity metric in question is output per employee, time saved on specific tasks, or revenue tied to AI-assisted work; whether the sample IBM studied resembles the reader's own industry and company size; and whether the $69.22 billion in projected market spending is concentrated in the same tool categories the report flags as underperforming. None of those answers are available in the aggregated coverage reviewed for this piece, which is itself the clearest signal that the underlying IBM report, not secondary summaries of it, is where the real analysis lives.