The 93% Gap: AI Investment ROI as a Blockchain Trust Problem

93% of business leaders cannot prove AI investment ROI. That number is a systemic vulnerability. Not in the code, but in the accounting. In DeFi, we’ve seen this pattern before. Liquidity is just trust with a price tag. Here, AI ROI is just trust with a missing measurement framework. The KPMG survey—a Big Four auditor’s report—drops a bomb: only 7% of executives can confirm positive returns on their AI spend. The rest are operating on faith. This is the same faith that fueled the 2020 DeFi summer without audits. Code is law, but bugs are reality. Similarly, AI investment is law, but ROI is a ghost function. Context: KPMG surveyed a cross-section of global business leaders. The headline is stark: 93% cannot prove their AI investments are generating returns. This is not a rejection of AI’s potential. It is a rejection of the current measurement infrastructure. The report lands at a critical juncture—enterprise AI budgets are shifting from 'exploration' to 'validation'. The 7% who can prove ROI likely have a rigorous measurement framework. The rest are flying blind. In my years auditing smart contracts, I’ve seen the same phenomenon: protocol teams deploy code without reentrancy guards, assuming the market will reward them. The market doesn’t reward trust—it rewards verifiable trust. The same applies to enterprise AI. Audit reports are promises, not guarantees. The KPMG data is a warning: the promise of AI value is not yet verified. Core: The problem is structural. AI’s impact is embedded in complex business processes, making it nearly impossible to isolate causality. When a customer service agent uses an AI copilot, how much of the improved resolution time is due to the AI, and how much is due to better training? No standard attribution method exists. This is the same attribution problem that plagues DeFi yield farming—is the yield from the protocol’s value creation or from token inflation? Yield is a function of risk, not just time. In AI, ROI is a function of measurement, not just spend. The 93% figure reveals that the majority of AI spending is defensive: companies invest to avoid being left behind, not because they have a clear ROI model. This is analogous to liquidity mining in DeFi, where protocols pay users to provide liquidity without a clear path to sustainable value. The result is a house of cards. When the market turns, these unbaked investments will be the first to collapse. Contrarian: The contrarian angle is that this problem is not AI’s failure—it is a failure of the measurement industry. The 7% who can prove ROI have likely built a custom measurement framework, often involving A/B testing, controlled experiments, and multivariate analysis. This is a hidden opportunity. The market for AI value measurement, observability, and FinOps tools is about to explode. But there is a catch: KPMG itself has a conflict of interest. As a consulting firm, it profits from selling 'AI value management' services. The report is a marketing move—a way to create demand for its own solutions. This is the same dynamic we see in the blockchain audit space: audit firms highlight vulnerabilities to sell audits. The underlying truth is that the 93% gap is real, but the solution is not a single consulting engagement. It requires a new trust layer—one that is transparent, immutable, and verifiable. This is where blockchain comes in. Takeaway: The next wave of value creation in crypto will not be DeFi or NFTs. It will be AI value verification as a service. Smart contracts can provide on-chain attestations of AI model performance, verifiable compute, and immutable audit trails for ROI claims. The 93% gap is a market signal: the demand for provable AI value is immense. The projects that solve this will capture the same trust premium that Chainlink captured for oracles. The question is: who will build the trust layer for AI ROI? In my experience, the teams that treat ROI as a smart contract—a set of deterministic, verifiable rules—will win. The rest will remain in the 93%.