Your $2.59T AI investment just produced 0.07% productivity growth. Here's the structural reason why.

Gartner projects $2.59 trillion in worldwide AI spending this year. A 47% jump.

US total factor productivity grew 0.07% over four quarters ending Q1 2026.

95% of enterprise generative AI pilots have produced no measurable effect on the bottom line.

The question isn't whether AI works. It does — at the individual level.

The question is why individual gains are destroying organizational output.

Meta's CTO reported code changes up 220% year over year. Features actually reaching users? Up 36%.

That gap is the story. AI makes one person hyperproductive. Then dumps the output onto everyone else.

More emails. More proposals. More code to review. More documents to read. More meetings to schedule.

One person's AI productivity is everyone else's information overload.

LinkedIn data shows applications per job posting roughly doubled since spring 2022. AI makes applying easy. It makes hiring managers drown.

MIT and USC researchers found AI-generated text in federal court filings went from 1% in 2023 to 18% in early 2026. AI makes filings faster. Judges now spend more time catching hallucinations than reading arguments.

Your organization isn't getting more productive. It's getting busier. Those are not the same thing.

Audit your AI deployments today. Measure organizational output, not individual throughput. If your teams are generating more but your business outcomes aren't moving, you've built an expensive hamster wheel.

SOURCE: https://www.computerworld.com/article/4215022/why-ai-doesnt-make-companies-more-productive.html
VERIFIED: Computerworld Aug 28, 2026; Gartner AI spending forecast; Fortune/MIT 95% pilot failure stat; St. Louis Fed productivity data
SIGNAL: The $2.59T AI spending number against 0.07% productivity growth is the single most important ratio in enterprise AI right now. This reframes the conversation from "AI doesn't work" to "AI works for individuals but destroys collective output" — which changes how you should deploy it.