Nearly half of financial services executives are struggling with managing credit risk as well as detecting and preventing fraud, with many now turning to AI to find a solution, according to a new study from risk-decision platform Provenir.

New findings from the firm’s 2025 Global Risk Decisioning Survey show that 49% of decision makers in financial services rate managing credit risks as a key challenge, alongside fraud detection (48%), developing risk decision processes (36%), and access to limited data (31%).

Over a quarter (26%) of the nearly 200 executives surveyed said that outdated technology was a significant barrier, with 59% reporting difficulty in deploying and maintaining their risk decisioning models.

Those issues are leading more financial firms to capitalise on AI tools for decision making, with 63% of respondents saying that they will invest in AI and embedded intelligence this year to help determine risks.

Over half (55%) of the CEOs, chief risk officers, and senior directors surveyed said that AI would help them to make streamlined strategy decisions and help to improve performances, while 53% see value in the ability to automatically tune AI models to make better, more accurate decisions.

Provenir’s results show that financial service executives also see a future where AI might replace the work of data scientists, with 45% saying that AI which could generate, deploy and monitor machine learning models for decision-making – without relying on the input of data experts – would provide them the most value.

Getting to that point might be tricky, however, seeing as 52% of execs cited the quality and availability of data as a  core obstacle to their plans, with a similar number concerned about the difficulty of integrating data sources into the decision process.

The survey found that streamlining data sources is a major issue of financial service firms, with 59% of respondents reporting that operating multiple decisioning systems is causing a lack of data flow and unified insights, and 52% saying it creates operational inefficiencies.

Recommended reading

• Report: AI Adoption Elevates Human-centric Skills
• Job Automation More Likely to Affect Women
• Will AI Enable Workers to Secretly Juggle Multiple Jobs?

When asked about data and fraud, 37% of executives said they struggle with effective data management for application fraud prevention, specifically not being able to easily integrate new data sources, while 36% reported that they face problems using current AI fraud solutions.

Nearly one-third (31%) agreed that the most important aspect for comprehensive fraud strategies is the ability to break down data silos between fraud and credit risk teams, while over a fifth (21%) cited challenges in gaining a comprehensive risk view of customer data.

“Financial institutions are keenly aware of today’s increasingly complex threat landscape and must adopt new approaches for improved risk decisioning and fraud prevention across the customer lifecycle while providing frictionless and personalized customer experiences,” said Carol Hamilton, chief product officer at Provenir. 

“With an AI decisioning platform more closely aligning credit and fraud risk teams, financial services executives can ensure holistic, end-to-end decisioning with a complete view of customers across the entire lifecycle.”