Esker survey: 72% of finance leaders overspent on AI budgets

Image
Esker survey: 72% of finance leaders overspent on AI budgets

LYON -- November 19, 2026 -- Seventy-two percent of finance leaders surveyed by Esker say their organizations spent more than planned on AI initiatives over the past year, according to the company's 2026 Global Finance AI Trust Index, published on September 24, 2026. Only 48% of respondents could cite a direct financial gain such as stronger cashflow, against 58% who cite productivity improvements, a gap of 10 percentage points by our calculation.

Esker, which sells AI automation software for finance departments, ran the survey in August 2026. It covered 338 finance leaders: 137 CFOs in the U.S., U.K., E.U., Australia and Canada, and 201 director-level and above leaders in the U.S. and U.K. The two groups differ in seniority and in geography, which limits how far the gaps between them can be read like for like.

CFOs report heavier cost pressure than other finance leaders

Of the CFOs surveyed, 75% report higher AI software and subscription costs, against 45% of other finance leaders, a 30-point gap. Some 65% of CFOs say they struggle to connect AI usage to specific business outcomes, versus 40% of their peers, a 25-point gap.

CFOs are also more positive on productivity: 70% say AI has improved it, against 50% of other leaders. Weighting those shares by group size (137 and 201 respondents) gives about 58%, which matches the overall productivity figure. Esker did not disclose how far spending exceeded plans.

CFOs rule out AI execution on revenue targets and headcount budgets

Across all respondents, 32% use AI to recommend actions that teams carry out manually, and 29% let AI act with human approval. Together that is 61%, by our calculation.

The line is firmer for strategic decisions. On setting revenue targets, 78% of CFOs say AI should not execute, compared with 11% of other leaders, a gap of 67 points and about 7 times the rate. On allocating headcount and hiring budgets, the split is 74% against 17%, or about 4 times.

Scott McDermott, Esker's own CFO, said CFOs "aren't rejecting greater AI autonomy" but are weighing the guardrails needed to raise their confidence.

Data quality and integration gaps hold back expansion for nearly half of leaders

Data quality is the most-cited barrier, named by 48% of respondents. Insufficient integration between finance systems and AI tools follows at 42%. CFOs are the driver of the second figure: 58% of them cite integration, versus 31% of other leaders, nearly twice the rate. Weighted by group size, those shares work out to about 42% overall, consistent with the headline number.

Governance exposure is wider than the barrier rankings suggest. Some 66% of leaders know or suspect that finance employees use unapproved AI tools, equivalent to roughly 223 of the 338 respondents by our calculation.

The report names three conditions for scaling AI in finance: measurable business value, reliable integration with core systems, and governance built into every decision and action. Esker says its Synergy Agentic Framework is designed around these conditions and covers source-to-pay and order-to-cash, tying greater AI autonomy to outcomes such as working capital and productivity.

Published by:
fairsonline_team
Company
Target market(s)
News Type