CFOs Face Pressure in Agentic AI Deployment: The Imbalance Between Speed and Governance
According to a survey commissioned by Avalara and conducted by Censuswide, the vast majority of CFOs (92%) face pressure to demonstrate ROI from agentic AI, with 71% indicating that the pressure primarily stems from deployment speed. However, only 7% of finance leaders believe their organizations prioritize governance over speed, and this figure drops to just 4% among U.S. respondents. The survey also found that 30% of respondents said their organizations had not updated internal controls in the past year to reflect the actions of AI agents, and 44% expressed only some confidence in explaining AI agent behavior to auditors or regulators.
If generative AI is a potential game-changer for the finance function, then agentic AI—which acts proactively rather than reactively—is even more disruptive. However, most CFOs say they feel pressured to deploy agents at a pace faster than what is practically feasible, and this pressure is overwhelming.
This conclusion comes from a June 2026 survey covering 1,505 CFOs and other senior finance leaders in the United States, the United Kingdom, Australia, and India, all of whom had deployed, piloted, or actively evaluated AI agents in the past 12 months.
The researchwas conducted by Censuswide and commissioned by Avalara. As a provider of agentic tax and compliance software, Avalara is a stakeholder, but the findings are still sobering.
The vast majority of respondents (92%) said they face significant (50%) or moderate (42%) pressure to demonstrate return on investment (ROI) from AI agents; of these, 71% said the pressure comes entirely or primarily from the pace of deployment.
The key issue: only 7% of finance leaders said their organization prioritizes governance over speed when it comes to agentic AI. Among U.S. respondents, that figure is just 4%.
30% of respondents said their organization has not updated internal controls in the past year to reflect actions taken or suggested by AI agents; 44% said they have only limited confidence in explaining AI agent behavior to auditors or regulators.
According to Avalara, these findings reveal a dilemma for the finance function: on one hand, executive pressure pushes for accelerated adoption of agentic AI; on the other, operational realities require careful management of agents, especially in tax and compliance, where decisions must withstand regulatory scrutiny.
"Speed without accountability creates new forms of risk, and speed without rethinking workflows limits ROI," said Hugo Sarrazin, CEO of Avalara, in the report.
Not only do CFOs lack confidence in understanding agentic AI, but 76% of respondents also said their organization lacks dedicated internal finance expertise to understand how agents work.
About a quarter of respondents (23%) said thataccountability for major AI errorswould be unclear or unassigned.
Meanwhile, although 38% of respondents said they have achieved "scaled" ROI from AI agents, 50% said returns are limited, and 12% said there is no clear return yet.