AI has gained confidence in the finance function without gaining a blank cheque. Deloitte’s latest UK CFO Survey finds 59 percent of finance chiefs are more optimistic about AI’s ability to improve organisational performance, up from 39 percent in the third quarter of 2024.
Technology investment also looks structural. Ninety-six percent expect UK companies to increase spending over the next five years, while 77 percent anticipate better productivity and business performance. Yet risk appetite remains only 15 percent, below the longer-term average of 25 percent.
Together, those numbers describe a demanding environment: boards expect digital investment to continue, but speculative projects still need to compete for capital against uncertainty, resilience, and immediate operational priorities.
AI Has Moved Into the Capital-Allocation Conversation
CFOs are no longer passive approvers of an IT budget. The survey positions them as stewards of technology strategy, particularly where AI promises process automation, stronger forecasting, and measurable productivity.
This changes how projects should be presented. A model benchmark or successful demonstration does not explain the financial result. Finance leaders need a baseline, expected benefit, full implementation cost, timeframe, operational owner, and evidence that the gain can continue after a pilot.
Tightly scoped uses are likely to fare better than open-ended experimentation. Automating a defined finance process or improving forecast accuracy can be compared with current performance. Broad promises to transform knowledge work are harder to govern, audit, and defend when budgets tighten.
Confidence Rises While the Wider Business Remains Cautious
Overall business confidence remains negative at net minus 13 percent. Only 17 percent of CFOs call capital expenditure a strong priority, slightly above its historical average. Investment is available, but programmes seen as poorly governed or disconnected from productivity remain vulnerable.
External risk reinforces that selectivity. Thirty-eight percent rate uncertainty as high or very high. Geopolitics is cited by 65 percent of respondents, while UK competitiveness and productivity reach a historically high risk rating of 62 percent.
AI business cases therefore compete with systems resilience, data security, energy efficiency, and supply-chain visibility. In practice, the strongest programmes may address several priorities together—for example, automating a process while improving auditability or using better forecasting to reduce operational exposure.
Technology Funding Now Comes With Human Conditions
Deloitte’s interpretation recognises that tools do not deliver productivity independently. Employees need training, workflows must change, information has to be reliable, and leaders must decide how automated output will be supervised.
These costs belong inside the investment case. A programme that funds licences but omits change management and governance understates what success requires. It may appear cheaper during approval and then struggle when adoption, integration, or control work arrives later.
Portfolio discipline matters too. If every department launches a separate assistant, duplicated subscriptions and inconsistent controls can consume the savings each project promises. CFO and technology leaders need a shared view of AI spend, dependencies, adoption, and benefits across the organisation.
Finance and technology teams need a shared measurement language. Productivity should specify whose time is saved, what happens to that capacity, and whether quality remains stable. Forecast improvements require comparison with previous error rates. Automation should include exceptions, incidents, and manual correction costs.
The survey is optimistic because CFOs increasingly see technology as necessary infrastructure for medium-term growth. It is cautious because capital still has alternatives and the external environment remains difficult.
For IT leaders, the opportunity is real but conditional. AI initiatives that translate technical capability into credible, auditable business value are likely to receive support. Those that rely on enthusiasm alone will meet a finance function that is more interested than before—and more prepared to ask exactly what the investment produced.
