UK finance leaders face pressure to rush AI agents

Avalara has published research suggesting UK finance leaders are under pressure to deploy AI agents faster than governance processes can keep pace. The findings are based on a survey of 505 UK chief financial officers and senior finance leaders.

More than nine in 10 respondents said they faced moderate or significant career pressure to show a return on investment from AI agent spending, with half describing that pressure as significant. At the same time, 54% said their AI agent initiatives had delivered only limited measurable return so far, while 74% said deployment pressure was focused mainly on speed.

The figures point to a gap between executive expectations and the controls needed for AI use in finance, where decisions can affect reporting, tax, compliance and audit processes. The report focuses on agentic AI, a category of systems designed to take actions or make recommendations within business workflows.

Governance weaknesses appeared across several measures in the UK sample. Among respondents, 77% lacked dedicated in-house finance expertise able to understand how their AI agents work, leaving many teams reliant on suppliers and IT departments.

Almost half, 47%, said they were only somewhat confident they could explain an AI agent’s actions to an auditor or regulator. Another 21% said accountability for a significant AI agent error in finance would either be unclear or rest with no one, while 48% said AI incident response plans were either untested or still being developed.

Control gaps

The survey suggests the issue is not resistance to AI adoption but uncertainty over how to supervise it once embedded in finance operations. Respondents said the most useful steps for raising confidence in wider deployment centred on trust, data quality and traceability.

Measures cited included AI agents operating within existing systems of record, outputs grounded in verified tax, compliance and financial data, validation against known compliance requirements, supplier commitments on accuracy and accountability, and audit trails documenting each AI action.

The two most valued functions were audit-ready documentation for every AI-driven action and monitoring regulatory changes with updates applied in real time. Those preferences suggest finance teams want tools that can withstand scrutiny rather than systems that simply move faster.

Avalara commissioned the study across four markets, surveying more than 1,500 chief financial officers and senior finance leaders in the UK, US, India and Australia. All respondents had deployed, piloted or actively evaluated AI agents in financial processes over the previous year and worked at companies with revenue above USD $10 million.

The international findings closely tracked the UK numbers. Across all markets, 92% said they felt moderate or significant career pressure to demonstrate AI return on investment, while half said their AI agent programmes had produced only limited measurable return to date.

Only 7% said their organisation prioritised governance over speed, and 30% said internal controls had not been updated within the past year to reflect AI agents taking or recommending actions. Another 44% said they were only somewhat confident they could explain an AI agent’s actions to an auditor or regulator.

Executive pressure

The research places finance leaders in the middle of a broader shift in corporate AI strategy. Many businesses now want AI systems to move beyond drafting text or analysing data into areas where they can initiate or recommend operational decisions.

That creates particular tension in finance because errors can be visible, difficult to reverse and subject to regulatory scrutiny. Tax calculations, reporting decisions and compliance steps often require a documented chain of accountability, something many organisations still appear to be building.

Hugo Sarrazin, Chief Executive Officer at Avalara, said the risk comes when adoption outpaces oversight.

“Finance leaders are right to move quickly to capitalize on agentic AI opportunities, but speed without accountability creates new forms of risk, and speed without rethinking workflows limits ROI. The organizations that realize the greatest value from AI won’t simply deploy more agents. They’ll leverage agents with trusted data, governed workflows, and clear controls that enable automation with confidence,” said Sarrazin.

External industry figures cited in the report made a similar point about the need for broader expertise. The challenge, they argued, is not only technical implementation but understanding what AI agents can access, what they can change and when human approval is needed.

“Finance leaders are being asked to move quickly with AI, but governing agents requires a new combination of domain, AI, IT, and data governance expertise. As AI agents gain access to financial and compliance workflows, organizations need to know what those agents can see, what they can do, and when human approval is required. That kind of control has to be built into the architecture, not added after the fact,” said Frank Cirone, VP Commercial Strategy at Snowflake, a cloud data platform company.

Jim Lundy, Founder, CEO and Lead Analyst at Aragon Research, framed the issue as one of explainability as much as automation.

“AI agents are now moving into business processes that require trust, transparency, and governance by design. As enterprises scale agentic AI, the question becomes less about whether the technology can act and more about whether organizations can understand, control, and explain those actions. In finance, where workflows are auditable and outcomes carry real business consequences, governance and explainability will become essential requirements for adoption,” said Lundy.