IrelandFinance and insurance2025 – ongoingBank of Ireland
Machine-learning card fraud scoring and AI-assisted contact centres
Bank of Ireland said AI assessed about one billion card transactions in 2025 and prevented €9.7 million in potential losses
Problem
The bank needed to detect card fraud while payments were in progress, across more transactions than staff could check by hand. Customers also spent time being transferred between contact-center departments.
Approach
The bank’s AI and machine-learning system scored each card transaction for fraud risk. The score determined whether a payment could proceed or needed extra verification. A team of 225 staff worked on fraud and financial crime. Separately, contact-center AI routed customers to the appropriate help and assisted staff with call analysis.
Results
Takeaway
Fraud scores guided decisions to approve payments or require extra verification at a scale staff could not review manually. The bank’s €9.7 million figure estimated potential customer losses prevented, rather than realized savings.
Categories
- Size
- Large enterprise
- Industry
- Finance and insurance
- Business function
- Finance, risk and legalCustomer service
- Innovation type
- Process innovation
- Adoption
- Built in-house
- Evidence
- Company's own claim
Sources
Source summaries
Bank of Ireland said AI and machine learning assessed one billion card transactions in 2025 and prevented €9.7 million in potential customer losses. Each transaction received a fraud-risk score that determined whether it could proceed. The bank said 225 staff worked on fraud and financial crime.
Bank of Ireland said AI support reduced contact-center call transfers by 40%. It also said it sent customers 127 million AI-generated spending analyses in 2025 and that about 20% of staff were taking or had completed AI or data-fluency training.
Full details are in the original sources.
