Research suggests there is some way to go for widespread AI adoption
Financial institutions worldwide are dialling back the hype around generative AI and putting more weight on responsible AI frameworks, according to new research.
The study from analytics firm FICO and Corinium Global Intelligence, surveyed more than 250 senior executives across banking and financial services, including chief analytics (CAOs) and AI officers (CAIOs), CTOs, and CIOs.
The findings point to a clear recalibration in AI priorities as firms seek measurable value and greater accountability from their deployments.
“Responsible AI extends beyond risk mitigation—it's a business imperative,” says Dr. Scott Zoldi, FICO’s chief analytics officer. “Over half of CAOs and CAIOs (56%) believe that implementing Responsible AI standards will significantly impact ROI. Meanwhile, human-AI collaboration is key, with 44% of surveyed leaders identifying it as an exciting area for future development. To ensure accountability and reduce AI hallucinations, organizations must clearly define the boundaries and interactions between human oversight and AI capabilities.”
The emphasis reflects growing recognition that responsible practices can unlock business value, not just manage risk.
The survey also found strong belief in the benefits of unified platforms with more than 75% of executives saying that greater alignment between business and IT leaders, along with consolidated AI infrastructure, could increase ROI by 50% or more.
Barbara Widholm, vice president of automation and AI at State Street, warned that fragmentation continues to undermine AI progress.
“From my perspective, the lack of a unified platform and process often leads to duplicated efforts, inconsistent tooling, and misaligned priorities. AI initiatives may be technically sound but fail to scale or integrate due to infrastructure gaps or unclear ownership,” she says.
Despite enthusiasm, most financial firms admit they are struggling to connect AI to broader corporate strategy with 95% of respondents saying that their AI initiatives are not consistently aligned with business objectives.
Only 12% reported they have fully integrated operational standards such as bias mitigation, performance monitoring, and data governance into their AI programs, a sign that many are still in the early stages of responsible adoption.
The report suggests that financial services is entering a new phase of AI maturity, but while generative AI continues to generate interest, the sector is increasingly prioritizing governance, explainability, and cross-team collaboration as foundations for long-term success.
