Bank of America executive urges CFOs to prioritize data before AI spending

Bank of America’s Matthew Davies said chief financial officers should define business outcomes and fix data foundations before committing to new AI, payments or treasury technology. In an interview with PYMNTS for its 2026 “Summer School” series, Davies said the larger threat for finance teams is misdirected spending rather than a failure to spend enough.

Davies, head of Global Payments Solutions for EMEA and global co-head of corporate sales for Global Transaction Services at Bank of America, said CFOs are being offered a widening set of tools, including artificial intelligence products, real-time payments capabilities, forecasting systems, treasury platforms and automation software. He said those investments should be measured against specific goals such as better liquidity visibility, stronger controls, greater efficiency or faster decision-making.

“The biggest risk and challenge is misinvestment rather than underinvestment,” Davies told PYMNTS. He said finance leaders should focus on business problems rather than adopting new technology because it is widely discussed.

Payments data becomes a finance priority

Davies said payments are moving beyond their traditional role as a back-office function and are increasingly linked to liquidity management, risk control and enterprise efficiency. The shift comes as companies manage cash across markets, currencies, banks and legal entities while dealing with geopolitical disruption, changing interest rates and fraud threats, according to PYMNTS.

Near-real-time information about cash positions can allow treasury teams to make funding and investment decisions more quickly, Davies said. It can also reduce reliance on fragmented reporting or end-of-day reconciliation when companies need to move liquidity across the business.

The wider value, Davies said, sits in the data attached to payments. Treasury teams are using payment information to support cash-flow forecasting, capital allocation and strategic planning, he told PYMNTS.

That makes integration and governance central to modernization. Payment infrastructure can generate real-time information for AI systems, while stronger data can improve forecasting and controls. Davies said those gains depend on companies connecting systems, setting rules for data use and securing adoption across teams.

Data quality comes before automation

Davies said fragmented information across enterprise resource planning systems, treasury platforms, bank portals and acquired businesses can limit the effectiveness of advanced technology. If data is not high quality and standardized, finance teams lack the base needed for automation, forecasting, financial decisions and AI, he told PYMNTS.

He added that many organizations find value in improving data quality before adding further technology infrastructure. For CFOs, that can change the order of investment: standardizing data, integrating systems and designing controls may come before more visible AI pilots.

Davies said the most immediate opportunity for finance teams is automating repetitive manual work across finance processes. That can include activities such as compiling reports, matching transactions and handling routine exceptions, according to PYMNTS’ description of the discussion.

Reducing those tasks can give finance staff more capacity for cash forecasting, scenario planning, risk assessment and strategic decision support, Davies said. He said AI should be applied where it addresses defined business challenges and produces measurable value.

Davies also said successful modernization requires cooperation across treasury, finance, technology, cybersecurity, data and risk teams. He called for change management, implementation support and phased rollouts that expand only after value has been shown.

For CFOs, Davies framed the 2026 technology test as one of discipline. The task, he told PYMNTS, is to choose investments that improve visibility, liquidity and decision-making while producing measurable business results.

This story draws on original reporting from PYMNTS.