AI Risk Accountability Splits Business and Tech Leaders: PwC Survey

What happened

New research from PwC's Digital Trust Insights survey finds that business and technology leaders remain divided over who should be accountable for managing AI-related risk inside their organisations. According to the findings, reported by ZDNET, there is no shared consensus on whether responsibility for AI oversight sits with technology leadership, risk and compliance functions, or business unit heads.

The research points to a growing debate over whether organisations need a dedicated AI leadership role — an "AI chief" or equivalent — to close this accountability gap, rather than leaving AI risk ownership distributed across existing functions.

Why it matters

As AI tools move from pilot projects into everyday business operations, unclear ownership of risk creates real exposure — in data handling, decision-making, compliance and customer-facing outcomes. When no single function is clearly accountable, issues can fall through organisational gaps until they surface as incidents, rather than being caught and managed proactively.

For leaders driving digital transformation, this is a governance and operating-model question as much as a technology one. Deciding who owns AI risk shapes how quickly and safely an organisation can scale AI use, and whether oversight keeps pace with adoption rather than trailing behind it.

The Renascence take

The debate over an "AI chief" often gets framed as an org-chart question, but the deeper issue is behavioural: ambiguity about ownership quietly signals to every team that AI risk is somebody else's problem.

Most organisations will default to bolting AI governance onto an existing function rather than genuinely redesigning accountability — and that's the mistake. Risk ownership that isn't tied to a clear, resourced mandate becomes theatre, not control. The operators who get ahead here won't just appoint a title; they'll define decision rights for AI use case by use case, so accountability is tested against real workflows rather than asserted on an org chart.