KPMG's Q3 2026 AI Pulse documents growing confidence in enterprise AI maturity. AIDE's public-evidence lens adds context on leadership capability, accountability and returns.
By the AIDE Institute
In brief
- The regional maturity gap narrowed from 16 points to 8, based on respondents' assessments of their organizations.
- "Established ROI" moved from 8% to 10%, while the largest shifts occurred in the early and middle stages.
- 89% report measurable value, while 12% consistently compare that value with cost.
Reported maturity is converging across regions
KPMG's Global AI Pulse Q3 2026 surveyed 2,131 senior leaders at organizations with $100M or more in revenue, across 20 countries, territories and jurisdictions, between July 23 and August 26, 2026. Respondents placed their own organizations on KPMG's maturity scale.
Share reporting their organization is scaling AI or beyond:
| Region | Q1 2026 | Q3 2026 |
|---|---|---|
| Americas | 56% | 64% |
| Asia Pacific | 43% | 61% |
| EMEA | 40% | 56% |
| Spread | 16 pts | 8 pts |
The shift indicates that leaders across regions increasingly describe their organizations in similar maturity terms. Observable evidence can help show how that convergence is taking shape in enterprise capability.
More organizations report progress beyond experimentation
Organizations describing themselves as still experimenting fell from 22% to 13%. Those reporting established returns moved from 8% to 10%.
KPMG describes a market that has "largely worked out deployment." The decline in organizations reporting that they remain in experimentation is consistent with broader adoption and greater executive confidence. A complementary evidence layer can help organizations assess how that progress appears in integrated systems, governance, usage and sustained returns.
Formal controls are becoming more common
55% of organizations report a formal AI harness layer. Among respondents reporting established returns, 86% have a cross-functional or enterprise-wide layer, including 65% enterprise-wide. Among experimenters, the figures are 31% and 10%.
The association between reported infrastructure and reported maturity is useful evidence of how organizations are structuring AI oversight. Because both measures are self-reported, operational evidence can add further insight into how those controls function and contribute to results over time.
Named ownership and leadership literacy
53% place responsibility for AI-informed decisions at C-suite level or above: 35% with a named executive, 18% with the CEO or executive committee. The remaining 47% have a non-executive owner or no clear one.
A named owner creates a clear line of accountability. Leadership literacy offers an additional view of whether decision-makers are equipped to challenge a model, interpret a trade-off and explain a decision.
The AIDE Institute's August 2026 public-evidence analysis found an observable AI-literacy signal for 240 of 413 uniquely matched S&P 500 CEOs, or 58.1%. Among 178 matched Matrix Quadrant (Sector) AI Trailblazers, the figure was 82.6%.
These are different cohorts and measures from KPMG's, and the AIDE signals are public role-level classifications rather than competency scores or proof of governance quality. Read together, the findings suggest that organizations can strengthen governance measurement by tracking named ownership and demonstrated literacy as distinct, complementary indicators.
Value measurement is advancing, with room for greater cost discipline
- 89% report some measurable value from AI.
- 10% classify themselves in the most advanced "Established ROI" stage.
- 12% consistently compare AI's value with its cost across the organization.
These findings show broad progress in identifying AI-related value and a clear next step in applying cost discipline consistently. Measuring materiality, attribution and full cost can help organizations translate early benefits into a more complete view of returns.
An additional board question
KPMG's three board questions provide a useful foundation: Who is accountable? How is value measured against cost? How will AI systems continue if a model, vendor or interface changes? AIDE's public-evidence lens adds a related question: What observable evidence shows how the management system is working?
Together, the two evidence approaches suggest a practical measurement agenda:
- Continue tracking respondent assessments of organizational maturity.
- Document cohort continuity so changes across waves can be interpreted with confidence.
- Pair named ownership with observable literacy and decision capability.
- Track measurable benefit alongside value assessed consistently against cost.
- Use operational evidence to examine how maturity and controls relate over time.
KPMG's findings show that leaders' descriptions of AI maturity are converging across regions.
Adding observable evidence on implementation, leadership capability and realized economics can provide a fuller view of that progress.
Sources
KPMG Global AI Pulse Q3 2026; AIDE Institute, CEO Literacy 2026. Methodology: The AIDE Index; Why the AIDE Index Doesn't Use Surveys.
