As finance teams begin to deploy AI agents, chief financial officers (CFOs) face a critical decision between focusing on immediate returns or prioritising governance and oversight.
The risks and opportunities of AI agents are fundamentally different from previous automation waves in finance.
Alex Levine, director analyst in the Gartner Finance practice, provides insight on how finance leaders should get started with AI agents.
What makes AI agents riskier than traditional automation, traditional AI or GenAI techniques in finance?
AI agents are riskier in finance because they act with greater autonomy. Unlike traditional automation, which follows fixed rules, or GenAI, which generates content, AI agents interpret objectives, plan and execute steps, and interact with multiple systems.
This autonomy means agents can make decisions and take actions that may not be visible or easily explained to humans. It also means errors or misjudgments can propagate quickly if oversight is unclear. Finally, the risk shifts from just the output to the entire process, making governance and traceability essential.
Why should CFOs treat their first finance AI agent as a governance pilot rather than a ROI pilot?
The greatest risks and the greatest value lie in establishing oversight, not in chasing quick financial gains. Early pilots are most likely to fail due to unclear controls, not poor technology.
By prioritising governance, CFOs can surface and address gaps in oversight, traceability, and review processes before scaling to use cases with higher stakes. This approach ensures that future AI deployments are built on a foundation of robust, auditable controls, protecting both the organisation and its stakeholders.
Success should be measured by governance readiness, not just autonomy or ROI.
Where should CFOs start when piloting their first finance AI agent?
CFOs should begin with a low risk, contained workflow where errors are visible and reversible. The ideal pilot:
- Involves a process with clear boundaries, repeatable steps, and verifiable outputs
- Avoids high-consequence processes, such as those requiring restatement or regulatory reporting
- Establishes agent boundaries, including data access, allowed actions, and required human review, before any development begins
- Assigns clear ownership across finance, IT, and audit/risk teams
- Operates in a sandboxed environment to observe and learn from agent behavior without operational exposure
This approach allows finance leaders to refine governance frameworks and build confidence before expanding to more critical areas.
How should CFOs measure whether a finance AI agent pilot is successful?
Success should be measured by governance readiness, not just autonomy or ROI. Key indicators of a successful agent pilot include practical and consistently applied controls and review points and achieving complete traceability of agent actions.
Another critical milestone is maintaining a comprehensive failure log that documents issues and fixes. The finance team should be able to reconstruct what the agent planned, accessed, and produced for every run.
Successful AI agent pilots should deliver templates and training that reduce the risk and cost of future deployments, as well as reusable governance capabilities that enable safe, scalable AI adoption.