Two years ago, AI was the CIO’s problem. Last year, it became the CFO’s budget question. This year, it has landed squarely on the COO’s desk, and the question is no longer just about deployment.
The shift makes operational sense. CIOs often own deployment, security, and governance. CFOs see the budget. But the COO is closest to whether AI changes how work gets done. COOs oversee the workflows where AI is supposed to create value, the teams being asked to work differently, and the operating metrics that show whether the change is working.
That puts COOs in a harder position. They’re being asked to produce evidence that the deployment investments showing up in the CFO’s budget are creating the operational improvements the board expects. Grant Thornton’s 2026 research found COOs are finding governance gaps in AI-affected operations that other C-suite leaders may not see quickly enough. Those gaps show up in cycle times, output quality, and workflow performance long before they show up in financial reporting. The accountability vacuum that forms when no one owns AI outcomes tends to be discovered by the COO first.
The old COO AI mandate was operational: get the tools deployed, train the teams, track adoption. The new mandate is to provide evidence: show that those tools changed how operations perform. Those are different measurement challenges.
Deployment metrics are easy. Did the tool go live? Are teams using it? Are adoption rates above threshold? Operational performance evidence is harder. Did cycle times change? Did output quality improve? Did workflow throughput increase? Did error rates fall? The COO who can answer those questions with data has a stronger position in every AI budget and governance conversation.
A COO needs measurement that connects AI usage to how work actually moves through the business. Four metrics matter most.
Cycle time is one of the clearest signals of whether AI is changing operations or adding a new tool to old workflows. The Workflow Intelligence platform captures workflow data continuously, making before-and-after comparisons possible without requiring manual measurement projects.
Department-level utilization numbers hide the variation COOs need to see. One team in a department may have 85% active AI utilization while another has 12%. That gap is an operational risk and a development opportunity, and it is invisible at the department level. The AI Adoption dashboard surfaces utilization at the team and role level.
High utilization without proficiency creates busy teams, not better operations. AI Fluency measurement tells the COO who is using AI and how well, which helps identify where proficiency investment would translate directly to operational improvement.
Ultimately, the COO’s AI evidence needs to connect to the metrics the CEO and board care about: output per employee, cost per unit, error rate, and customer-facing quality measures. The AI Impact platform connects AI usage to the business outcomes it is supposed to produce.
Grant Thornton’s finding that COOs are discovering AI governance gaps other C-suite leaders may not see reflects a structural reality: operational dysfunction shows up in workflow performance before it shows up in financial data. The AI governance adoption gap that 77% of organizations are experiencing is felt most directly by the COO: teams using ungoverned tools, workflows with unmonitored agent activity, and performance data that doesn’t match the adoption metrics IT is reporting.
Workflow cycle time changes, AI utilization by team and role, proficiency distribution across operational functions, and outcome attribution connecting AI usage to specific business performance metrics. Together, those measures tell a complete operational story rather than just counting how many employees have access to AI tools.
The CIO is primarily accountable for deployment, security, and governance. The COO is accountable for operational performance. AI is valuable to the COO when it changes how operations perform, not when it is merely deployed. That shifts the measurement focus from adoption metrics to outcome evidence.
Because operational dysfunction is often the first place governance gaps become visible. A team using an ungoverned AI tool may not create a financial reporting problem immediately. It creates a workflow performance anomaly that shows up in operational data before it shows up anywhere else.
The COO often has the clearest line of sight into whether AI is changing operational performance. That makes the COO a natural anchor for AI outcome accountability in operations. The broader question of who owns AI outcomes across the enterprise still requires input from the COO alongside the CIO and CFO.
Larridin’s Workflow Intelligence and AI Impact platforms give COOs the operational visibility they need: real-time data on how AI is changing cycle times, workflow quality, and output across every department, not just technology.
Book a discovery call to see the operational AI picture.