Rising application of AI in EU banking and payments sector
25/09/2025
Over the past decade, the EU banking sector has undergone a profound digital transformation, embracing a broad
spectrum of technologies to enhance operational efficiency and customer experience. Among those technologies, AI
has become increasingly prevalent and is playing a pivotal role in reshaping banking processes.
AI refers to a broad range of machine-based systems designed to operate with varying
levels of autonomy and which generally adapts to input data and context.
Observed use cases and market trends
The EBA monitors the adoption of AI, including relevant use cases and market trends. To-date, the EBA observes
92% of EU banks are currently deploying AI, and 8% are pilot testing or discussing AI use cases.
Profiling or clustering of clients or transactions AI enables the analysis of very
Delineating customers profiles according to their behaviour, large and unstructured datasets
preferences or transaction/credit history and the identification of non-
obvious clusters and patterns.
Grouping customers according to their similarities
Optimisation of internal processes
AI enables automatisation of
Summarising and classifying documents tasks in a way that improves
Preparing meeting minutes time- and cost-efficiency.
Improving IT applications, including generating codes
Creditworthiness assessment and credit scoring AI improves accuracy and
Evaluating the creditworthiness of individuals predictive power thanks to a fast
Assigning credit scores to individuals analysis of complex and vast
amounts of data.
AML/CFT and Fraud detection
AI enables a faster,
User identification and verification, including remote
more accurate, and scalable
onboarding and digital identification analysis that improves speed and
Detection of patterns and anomalies related to ML/TF and efficiency in the identification,
predicate offences, also for enhanced risk profiling verification and detection of fraud
Fraud detection or suspicious activities.
Real-time monitoring of user activity and transactions
Risk modelling
Detecting anomalies in transactions patterns (amounts, AI enables an
frequencies, counterparties) that may indicate fraud, improved detection of abnormal
operational errors or risks results, flaws, market trends,
shifts in customers’ preferences.
Analysing customer sentiment to indicate cases that warrant
further investigation or risk mitigation
Customer support
AI improves the
Customer support, including chatbots personalisation and responsiveness of
Other customer-facing applications customer service channels and of
digital customer experience.
Use of General-Purpose AI (GPAI) and Agentic AI in consumer-facing applications
The EBA monitors the adoption of GPAI and so-called Agentic AI in the EU banking sector, with a
focus on consumer-facing applications. The EBA is engaging directly with market participants,
consumer organisations and supervisors.
55% of surveyed banks are already using GPAI or agentic AI
in consumer-facing processes. Most common uses:
detection and notification of fraudulent or suspicious
activities,
assisting customer service agents and call centre
operators to handle service requests,
automating the provision of information or guidance
for customers to self-serve for digital actions,
automating support and financial education tools (e.g.
FAQs or Q&As), and
making digital assistants, including voicebot assistants,
available for banking customers.
Relevant use cases of GPAI and AI agents are:
summarising & obtaining drafting legal, support or
coding & programming
insights from documentation marketing documents,
The adoption of GPAI & agentic AI is ultimately subject to balancing of risks & opportunities
reliance on third-parties
Drivers improvements in productivity and Challenges quality of input data, adequate data
efficiency of staff governance & human oversight
optimisation of processes ensuring consumer consent to use data in
ability to enhance customer training GPAI models
interactions consumer access to clear & meaningful
explanations on AI logic & risks
reputational and litigation risks as a result
of hallucinations or inappropriate /
inaccurate information
