Only 7% of Executives Can Show AI ROI as Enterprise Agent Costs Mount
Only 7% of executives surveyed by KPMG said their organizations had established measurable ROI from AI, while nearly half had questioned, delayed, or reduced AI-agent deployments after costs exceeded expected benefits.
Why it matters: The story quantifies the gap between AI deployment and measurable enterprise value, making cost governance and ROI measurement central to the next phase of AI adoption.
Companies are deploying artificial intelligence faster than they are proving that the technology generates measurable financial returns. Only 7% of executives surveyed by KPMG said their organizations had established return on investment from AI, even as adoption of artificial-intelligence agents continues to expand. Reuters reported that nearly half of executives surveyed by KPMG had questioned, delayed or scaled back deployments of AI agents after costs exceeded expected benefits, while a separate BCG survey found that 94% of executives still planned to continue investing in AI. KPMG also found that organizations with strong visibility into AI operating costs were five times more likely to report established ROI than organizations without that visibility. AI agents can generate costs through inference, model usage, software subscriptions, integration work, security controls and autonomous workloads that consume computing resources continuously. The next phase of enterprise AI competition may depend less on which organization deploys the most agents and more on which can measure what those agents cost, what they produce and whether the economics scale.