ReportsFifth Annual Enterprise AI Report 2026
Report Overview
The Fifth Annual Enterprise AI Report from Domino Data Lab is one of the most direct assessments of the ROI gap in enterprise AI available. Conducted by BARC Research on behalf of Domino Data Lab, the survey covered 639 senior enterprise AI leaders - Director level and above - at organizations with annual revenues of $100 million or greater, across North America, the UK, and continental Europe. It focuses specifically on regulated industries: financial services, life sciences, and the public sector - the environments where AI governance and measurable outcomes matter most.
Key Findings
On the ROI plateau
57% of enterprises' AI ROI fails to outpace their investment - a number that has been unchanged since 2025 despite significant increases in AI production capability. The plateau is not a capability problem. It is a design and integration problem.
On production capability
93% of enterprise AI leaders report improved AI production capability in 2026 - up from 88% in 2025. Organizations can build AI faster than ever. They are still not converting that capability into financial returns at the same rate.
On the MIT NANDA parallel finding
The report cites independent MIT NANDA research reaching the same conclusion from a different methodology: 95% of organizations running AI pilots saw no measurable P&L impact. Only 5% of integrated pilots generated meaningful financial value. Two independent research methodologies. The same answer.
On regulated industries specifically
The report focuses on financial services, life sciences, and the public sector - environments where the gap between AI capability and AI governance is most consequential. In these sectors, an AI system that produces an output the organization cannot audit, explain, or defend is not a solution. It is a liability.
Why This Matters for Enterprise Leaders
The persistence of the ROI plateau - unchanged at 57% despite two years of improving production capability - is the most important finding in this report. It tells you that the barrier to AI ROI is not technical. Organizations can build AI faster and better than ever. They are still not making money from it at the same rate.
The MIT NANDA finding sitting alongside Domino's own research is particularly significant. When two independent methodologies - a large survey of enterprise AI leaders and an academic analysis of 300 public AI initiatives - arrive at the same conclusion, the conclusion is reliable. 95% of AI investments are not producing measurable P&L impact. The 5% that are have something in common - and it is not which model they chose.
Anterra's Take
What separates the 5% from the 95% is consistent across every research methodology that has studied it: the organizations getting real returns from AI designed their solutions around specific business problems that AI was the right answer to - and built them into the operation rather than on top of it.
The regulated industry focus of this report is directly relevant to the clients we work with. Healthcare, financial services, government - these are environments where AI governance is not optional, where the audit trail matters, and where an AI system that cannot be explained is worse than no AI system at all. We design for that standard from the first architectural decision.
Source: Domino Data Lab / BARC Research, Fifth Annual Enterprise AI Report 2026. Survey of 639 senior enterprise AI leaders, April 2026.
More reports
Ready to learn more?
Speak with one of our experts about what this means for your operation.
Let's talk