When Everyone Owns AI, Who Is Actually Accountable?

What Is Changing

In a 2026 survey conducted with EY, HFS Research asked 302 Global 2000 executives about obstacles to AI-related organizational change. Forty-two percent named siloed functional ownership as a structural barrier. Separately, 21% identified decision rights embedded in legacy roles as a work-design barrier.

These are executives' assessments of operating-model obstacles. The percentages do not show how many individual AI uses lack a named accountable owner. They do reveal why leadership should test whether shared AI work has a clear path for cross-functional decisions.

Why It Matters

Consider an illustrative customer-retention program. Marketing selects the offer, technology maintains the AI service, finance sponsors the investment, and customer operations manages complaints. Risk reviews exposure. Each function has a defined contribution.

Now an offer reaches a customer group excluded by the intended eligibility rules. Marketing wants to continue while it reviews the campaign. Customer operations recommends pausing the offer. Risk proposes new conditions. Finance asks how the decision affects the expected return.

The unresolved question is who can approve a pause, a customer remedy or revised eligibility rules. If the approved delegation is silent, the enterprise needs an escalation route to the designated decision authority. Naming an AI committee or a technical owner does not, by itself, settle that conflict.

The 6xD Reading

NIST's AI Risk Management Framework calls for clear risk-management roles and communication lines (GOVERN 2.1), with executive leadership responsible for decisions about AI-related risks (GOVERN 2.3). The OECD AI Principles address accountability and traceability according to each actor's role. Neither source prescribes a single business owner for every AI initiative.

The 6xD Enterprise AI Governance System applies those governance concerns through one accountability evidence record for each material AI use. The record connects decision authority, investment sponsorship and business outcome ownership, together with approval limits, escalation rights, measures and review dates. These are distinct responsibilities, although suitably delegated roles may overlap.

Through D2, Digital Cognitive Organization, the business executive answerable for the affected process stays visible alongside the technology owner. D4, Digital Transformation 2.0, connects the decision to the investment case and the result management expects to achieve.

Leadership Takeaway

At the next AI portfolio review, select one use shared across several functions. Present a realistic dispute: marketing wants to continue, customer operations wants to pause, and risk requires changes. Ask management to show who is authorized to decide, where the issue escalates if limits are exceeded, and which executive owns the resulting customer or business measure.

If those answers depend on an informal agreement, assign a corrective owner and deadline. Resolve any material authority gap through the existing governance process before extending the use.

See how to connect AI decision authority, investment and outcome ownership →