AI returns are rising faster than AI operating discipline
Deloitte reports more organizations attributing revenue growth to AI, but many still cannot see which activities create cost or value. For CHROs and CLOs, that is a governance problem before it is a scaling story.
Deloitte says 28 percent of surveyed organizations now attribute revenue growth to AI investments, up from 20 percent earlier in 2026. Deloitte also says many companies still lack detailed insight into which AI activities are driving costs and benefits. That is progress, but it is not yet a clean basis for bigger capability spend.
I would hold back broad AI training or role redesign justified by portfolio level return claims. Revenue growth tied to AI does not tell a CHRO, CLO or transformation leader which work changed, where judgment still sits, or which manager, team or workflow owner is accountable for the result.
The costly mistake usually comes next. L&D builds programs, talent teams adjust profiles, and business leaders push wider adoption before anyone has shown which tasks improved performance and which controls protected quality. Then capability investment gets attached to access and enthusiasm instead of changed work. An organization may be able to point to a gain and still be unable to show what to repeat, what to stop, or what to govern more tightly.
Deloitte's report does not provide detailed case evidence or long term ROI proof. It also does not show which governance choices separate stronger performers from the rest or establish a market wide operating model shift. So the decision test should stay narrow: require evidence that each material AI use case has a workflow owner, a defined performance change, a review control, and a way to measure cost and benefit after adoption. Without that, the next round of spend is partly a bet that visibility will arrive later.
Some organizations do report satisfactory ROI within two to four years. So the issue is not whether AI can produce value. It is that returns can show up before the organization knows which work design and controls produced them. Until returns are tied to changed work and accountable owners, the burden of proof should stay with the executive asking to scale adoption, training or redesign.
The signal I’m watching
Whether organizations begin reporting AI returns with use case level attribution, named workflow owners and explicit review controls, rather than portfolio level claims alone.
What would strengthen this signal
Comparable evidence showing that organizations with clearer use case governance can sustain better AI returns over time, not just report them once.
Sources
- Axios — Axios(2026-11-03)
- Deloitte Global — Deloitte(2025-10-22)
- Deloitte Insights — Deloitte(2023-10-27)
- Deloitte Survey — Deloitte(2026-10-29)