Your Board May Be Reviewing AI Activity Instead of AI Value

What Is Changing

McKinsey's 2026 State of AI survey found that 80% of respondents said AI improved their own productivity, while 37% reported a positive contribution to their organization's earnings before interest and taxes (EBIT). These are self-reported measures of different outcomes, not audited returns (McKinsey & Company, 2026).

Individual productivity gains may be real without yet producing a measurable enterprise-wide financial effect. Some benefits are nonfinancial, such as better service or lower risk. Directors still need evidence for the specific business result management promised.

Why It Matters

Imagine an AI assistant supporting insurance claims. A dashboard lists its launch date, adoption and AI-assisted claims. It does not show whether resolution became faster, rework fell or customers received better service.

The 6xD Reading

From the 6xD Digital Cognitive Organization perspective, the report should connect AI use to the work and its outcome. Management should show the baseline, what changed in claims handling and the observed result. A named business executive should explain any shortfall.

Averages can conceal a different result. A shorter claims-resolution time may reflect a shift toward simpler cases, while complex claims remain unresolved. A credible board report should compare similar claims over a defined period, identify meaningful changes in staffing or process design and show whether rework and repeat customer contacts also changed.

That comparison helps assess AI's plausible contribution but cannot isolate causation on its own. Management should report added costs, errors and complaints alongside the improvement.

Leadership Takeaway

At the next board review, select one material AI use. Request a short record showing adoption, the workflow change and the outcome against its baseline. Ask what evidence supports AI's contribution, what else may explain the result and when management will review it. Keep forecast benefits separate from achieved results.

Where the intended outcome is not yet measurable, management should report it as pending, explain the reason and set a review date. The board challenges the evidence and exceptions; management owns delivery and corrective action.

Board test: If AI use doubled next quarter, what would prove that the enterprise is better off, and who would explain a result that fell short?

Assess whether your board can govern AI value beyond adoption →