AI is changing more than how companies work. It is changing how risk enters an organization. For Donovan Hawse, Chief Executive Officer and Co-Founder of ACCIPRA, this requires finance leaders to rethink how enterprise risk is governed. “The central theme here is not whether companies use AI. It’s whether the company can trust and take responsibility for the decision that AI helps them make.”
After more than 25 years leading finance, audit, and risk functions at enterprise scale, Hawse sees a clear distinction between the fundamentals of risk management and the way those fundamentals must now be applied. The principles remain familiar, but there must be a new way of doing business.
AI Has Changed the Speed and Scale of Risk
Traditional risk management often relied on established systems, processes, and decision points where risks could be identified and assessed. AI introduces more entry points. Risk can emerge through software, employee experimentation with tools, workflows or third-party data. Once deployed, AI systems can also evolve as models learn and underlying data changes. The result is a risk environment in which problems can move faster than traditional controls are designed to handle.
“AI has changed the speed and the scale of risk,” Hawse says. AI-generated outputs can influence thousands of decisions before an issue becomes visible, making periodic assessments and static control reviews increasingly inadequate. Accountability, risk assessment, effective controls, testing, and escalation remain essential. What changes is the cadence. Organizations need continuous monitoring, stronger links between data and controls, and clear human oversight that allows AI applications to be evaluated, validated or suspended when outputs are unexpected or unexplained.
Finance Has a Unique Role in AI Governance
Finance leaders are particularly well-positioned to take on this challenge because finance touches nearly every part of an organization. Capital allocation, cash flow, performance, technology, reporting, and accountability all intersect through finance. That enterprise-wide perspective gives chief financial officers and other finance leaders a valuable lens for understanding where AI is influencing decisions and whether those decisions ultimately affect revenue, costs, cash flow, or enterprise value.
Hawse argues that finance leaders should be asking fundamental questions:
• “What decision is AI supporting?”
• “What data is it using?”
• “Who owns the outcome?”
• “How is performance measured?”
• “What happens when something goes wrong?”
Those questions turn AI governance from a technical exercise into an enterprise discipline.
Visibility Comes Before Governance
One of the most immediate challenges is simply knowing where AI is being used. “You cannot govern AI applications that you don’t know exist,” Hawse says. The rise of “shadow AI,” much like the earlier emergence of shadow IT, makes visibility particularly important. Employees can introduce AI tools into workflows without centralized approval, potentially creating exposure around data, decision-making, and accountability.
An effective governance model, therefore, begins with an enterprise inventory of AI use cases. Each application should have a business owner, an appropriate risk classification, approved data sources, and defined human oversight requirements. From there, governance needs to be embedded into the processes where AI actually has influence, including financial reporting, pricing, customer decisions, and hiring. This is where intelligent delivery and AI-augmented operations require more than technological adoption. They require scalable systems that connect technology to ownership and control.
The Biggest Risk Is Reliance Without Accountability
Hawse identifies one risk above others: unmanaged reliance on AI-generated decisions. The problem is not simply that an AI system might produce an incorrect answer. The greater danger comes when an organization embeds that answer into forecasting, reporting, pricing, customer operations, or other material workflows without sufficient validation.
For finance leaders, this creates a direct connection between AI governance and enterprise value. The question is no longer whether a particular tool is accurate in isolation. It is whether the organization understands where AI influences material decisions, how those outputs are validated, and who remains accountable for the outcome. That is ultimately where finance discipline meets AI. Strong controls can create the foundation for organizations to use AI with greater confidence, while preserving human judgment and transparency.
For Hawse, this is part of a broader argument about rebuilding enterprise operations for the modern era. Scaling does not necessarily mean building bigger teams, but building better systems, with the right combination of technology, human judgment, accountability, and operational discipline.
Follow Donovan Hawse on LinkedIn or visit his website.
