CFOs lack data trust, stalling AI adoption in finance

CFO Recruit has warned that finance leaders risk stalling fintech and AI adoption because they do not trust the data in their own systems. Ahead of World Fintech Day, the firm released research highlighting a widening gap between AI ambition and execution inside finance teams.

For its 2026 report, the executive search firm surveyed more than 500 Chief Financial Officers at North American companies with more than USD $50 million in revenue. The findings show that while most finance chiefs now discuss AI, many still lack a clear adoption plan and continue to work with fragmented finance data.

Christine Schneider, Regional Director (North America) at CFO Recruit, said this tension is shaping how boards hire senior finance leaders. As organisations prepare the next round of finance technology budgets and consider automation, AI and data tools, some still rely on finance platforms that cannot provide complete, accurate data.

"When we surveyed more than 500 CFOs at North American companies with more than $50 million in revenue for our 2026 report, 90% told us AI was on their agenda, but 70% have no defined plan to put it into practice. As the sector marks World FinTech Day, that gap between interest and action shapes much of what I see in the hiring market because boards want finance leaders who can close it. The obstacle has less to do with the tools than with the data behind them. Many finance functions cannot provide clean, consistent data from the systems they already use, so a forecasting model returns a faster version of a problem nobody has fixed. That helps explain why 82% of our clients put process automation ahead of any AI rollout.

"The same gap creates pressure across the whole finance team. Junior accountants already use AI for expense reports and forecasting in companies with no AI policy and open questions around data integrity. At the same time, CFOs move more carefully because they are responsible for the risks that come with unreliable data, security concerns, changing regulations and board questions about return on investment.

"External research points to the same concern. Gartner found that 59% of finance leaders used AI in their finance function in 2025, compared with 58% a year earlier. It also expects organisations to cancel more than 40% of agentic AI projects by the end of 2027 because of rising costs and unclear business value. Adoption has slowed as companies face the harder work of improving data quality and governance.

"The CFOs we place this year will lead finance teams for the next decade, and the ones who understand where finance technology adds value and where it does not will shape how their organisations spend the next round of investment."

Schneider's comments point to a two-speed shift inside finance teams. Junior staff often experiment with AI tools for tasks such as expenses or forecasting, while senior leaders weigh regulatory, data and investment risks before committing to wider deployment.

The research suggests boards now screen CFO candidates for experience in process redesign and automation. Many want leaders who can restructure workflows, standardise data and simplify finance operations before funding AI pilots at scale.

Vendors in finance operations report the same pattern. Businesses may have digitised core workflows, yet still rely on manual intervention for complex or exception-heavy tasks.

Caitlin Leksana, Co-Founder and Chief Executive Officer of accounts receivable automation firm Fazeshift, said many companies hit a ceiling with their current tools once they move beyond straightforward transactions.

"Most AR software gets a team to maybe 60-80% automation, then it stalls. The last 20-40% sits in an exception queue because no two customers pay the same way: different portals, different paperwork, different person you have to track down. Nobody has really gone after that part. Our agents plug into the ERP, the bank, email, the CRM - whatever is already running - so they do not just flag a late invoice. They figure out why it is late: a pricing dispute, a bad contact or a form that never got submitted. Then they handle it.

"One customer had our agents work through 9,000 payment communications in a single day. Across our customer base, we are automating more than 90% of manual AR tasks. CFOs do not need a better dashboard telling them what is wrong. They need fewer things going wrong, and a team that spends its day making calls instead of chasing emails."

Her comments underline a shift in focus from analytics to direct intervention in finance workflows. Vendors are building agent-style systems that sit on top of existing finance platforms and communications tools, then act on the information they uncover rather than simply surfacing alerts.

Data quality and identity assurance remain central concerns as finance organisations layer these systems onto customer-facing channels. Fraud trends have pushed banks and fintechs to extend checks beyond initial onboarding into every interaction that carries financial or privacy risk.

Clive Bourke, President of EMEA and APAC at Identity Specialist Daon, said institutions now need to treat identity as a continuous process rather than a single event.

"World FinTech Day is an opportunity to recognise how innovation has made financial services more accessible and convenient. But none of that works without trust. A bank or fintech needs confidence that it is dealing with the right person at every important point in the customer relationship, not only when an account is opened.

"Passing an identity check during onboarding is no longer enough. Financial institutions need to take an end-to-end view of identity, from initial verification through authentication, account recovery and transaction approval. If someone accesses an account, changes personal details or moves money, the institution must be able to assess the identity, the context and the level of risk at that moment.

"As customers move between devices and channels, financial institutions need to maintain confidence that the same person is behind each interaction. That is becoming harder as fraudsters use AI to imitate a customer's face or voice. AI agents add a different identity question. If an agent is acting on a customer's behalf, the institution needs to know who authorised it, what it is allowed to do and when the customer needs to step back in and approve an action.

"The principle behind continuous identity assurance is straightforward: keep low-risk interactions simple and introduce stronger authentication when the risk changes. You can balance security and convenience, but only if you continue to establish who is acting throughout the customer relationship."