IT 
ES 
A chat with our expert: Gaetano Ziri, Innovation Manager at Auriga
Artificial intelligence has moved beyond the realm of future-gazing in financial services and is quickly becoming a fundamental layer within modern banking architectures.
To explore the benefits of AI in banking and how this technology is reshaping customer experience, boosting operational efficiency, and influencing the work of technology providers, we spoke with Auriga’s Innovation Manager and Software Engineer, Gaetano Ziri, who shares his perspective on the most significant opportunities and challenges emerging across the sector.
What’s the biggest misconception people have about AI in banking?
A common misconception is that AI is a “plug-and-play” capability, a single component that can be dropped into existing systems to produce immediate transformation. In practice, AI requires strong data foundations, integrated channels, continuous model training, and robust governance.
Moreover, AI is not intended to replace human decision-making. Instead, banks are adopting augmented intelligence models in which AI supports processes, enhances analysis, and automates repetitive tasks, while humans continue to oversee complex judgment and strategy.
How do you see AI changing the everyday customer experience in banking?
AI is going to help shift the banking experience from reactive to proactive. Instead of waiting for customers to initiate interactions, banks will be able to anticipate needs based on behaviour, context, and historical patterns. This will support more relevant product recommendations, timely notifications, and streamlined digital journeys.
When it comes to physical and remote channels, AI-enabled analytics and orchestration will make it easier for banks to offer consistent and personalised experiences and customer journeys across mobile, internet banking, branch, and self-service terminals.
What do you think banks are still getting wrong about leveraging customer data, and how might AI help unlock its true value responsibly?
A lot of institutions still manage their customer data in silos, and transactional, operational and behavioural data often remain disconnected. This fragmentation reduces the quality of insights and data analytics in banking that AI models can generate.
You get the most value from AI when your data is unified, governed, and enriched. Responsible AI frameworks, supported by explainability, consent-driven usage policies, and transparency, allow banks to extract insight while maintaining trust.
How is AI helping banks create smoother, more personalised services across channels, and how does this apply to the ATM and self-service channel?
AI-driven analytics enable banks to manage journeys across channels more intelligently, supporting a seamless transition between mobile, branch and self-service environments.
In the self-service channel, including ATMs, AI mainly enhances operational performance and security rather than the user interface itself. Examples include:
- Improved user experience, through more consistent and context-based service delivery, supported by omnichannel platforms.
- Enhanced security, where AI-supported threat analytics and Zero Trust principles help detect anomalies and protect critical ATM infrastructure.
- Better channel integration, which ensures the continuity of the service between self-service terminals and digital channels.
How is Auriga applying AI within its omnichannel banking solutions to enhance operational efficiency, optimise cash management, and improve both customer and employee experiences?
We apply AI in a targeted and pragmatic way, focusing on areas where data-driven prediction delivers immediate operational value.
The primary and explicitly documented use of AI within the WinWebServer (WWS) suite is in cash forecasting and optimisation, a capability embedded in WWS Cash Management.
While within WWS Cash Management AI driven cash-flow forecasting is applied specifically, the broader WWS platform leverages advanced predictive analytics in banking, automated monitoring, and centralised orchestration to support:
- Proactive maintenance and improved uptime (WWS Proactive Monitoring OMNIA)
- Personalised services and marketing orchestration (WWS One-to-One OMNIA)
- Enhanced customer interaction through the NextGenBranch model, which streamlines service delivery through remote assistance and omnichannel continuity
This approach allows us to combine AI in areas where it drives the most measurable impact, notably cash forecasting, complemented by data-centric automation, across the rest of its omnichannel ecosystem.
What are some challenges banks face when trying to bring AI into their systems?
Banks still face several structural obstacles when trying to scale AI across their organisations. Many institutions continue to operate on legacy infrastructures that make real-time data access difficult, while information is often fragmented across systems and lacks the consistency needed for advanced analytics.
Integrating AI into multiple service channels, mobile, online, branch and self-service, adds further complexity, especially where architectures were not designed for seamless orchestration.
At the same time, banks report persistent skills gaps in areas such as data engineering, AI governance and Machine Learning Operations, which slows down deployment and long-term maintenance.
These challenges are compounded by evolving regulatory expectations around transparency and explainability.
Ultimately, successful AI adoption requires not only a solid technological foundation but also organisational readiness, clear governance frameworks and continuous cross-functional collaboration.
A recent European Banking Authority (EBA) analysis shows that although many European banks are experimenting with AI, only a portion have moved beyond pilots.
What’s holding banks back from full AI integration?
According to the European Banking Authority (EBA) only around 40% of EU institutions have moved from experimentation to actual implementation of generative or advanced AI models. Most projects are still in pilot mode or the early-development phases.
The main barriers are trust, scalability, and governance. Banks often succeed with limited pilots but struggle to industrialise AI due to fragmented data pipelines, inconsistent model-monitoring practices, legacy infrastructure, and the need for closer coordination between business, technology and risk teams.
Regulatory uncertainty also contributes to caution, especially for high-impact applications such as credit decisioning, fraud detection or customer-interaction automation, where explainability and accountability requirements are becoming stricter.
Are smaller banks and fintechs keeping up with big banks when it comes to adoption of AI in banking?
Fintechs and smaller banks are often faster in experimenting with new technologies thanks to more flexible, cloud-native architectures and fewer legacy constraints. However, data shows that large banks still maintain a structural advantage when it comes to scaling AI in financial services.
According to Bain & Company’s 2024 financial services survey, institutions with larger customer bases and richer datasets achieve up to 30% higher productivity gains from AI-driven initiatives, particularly in areas such as risk analytics and operations.
At the same time, the European Banking Authority reports that AI deployment remains uneven across the EU, with smaller institutions more likely to remain in pilot stages. This does not mean that smaller players cannot compete.
Modular and vendor-agnostic platforms such as our WWS suite allow institutions of any size to adopt advanced capabilities progressively, without disrupting operations. The real differentiator is less about size and more about the ability to modernise infrastructure, govern data effectively and drive cross-functional innovation.
Do you see challenges or ethical concerns when banks rely on AI for decision-making?
Absolutely. As AI becomes more deeply embedded in financial processes, banks must manage a new layer of ethical and operational risk. Models used for credit scoring, fraud detection or customer segmentation can inadvertently reproduce biases present in historical data if governance is weak.
Highly complex algorithms may also reduce transparency, making it harder for institutions to explain decisions to regulators or customers, an issue that is becoming increasingly important as supervisory bodies tighten expectations around accountability.
There is also the broader concern of over-automation, where critical contextual elements might be overlooked if decisions rely too heavily on machine outputs. In addition, the use of behavioural or biometric data raises questions about consent, proportionality and long-term storage.
Managing these risks requires a combination of robust oversight, explainable modelling techniques and clear communication with customers on how AI adoption in banking influences their financial interactions.
As AI becomes central to banking operations, how do you expect regulatory frameworks and consumer expectations to evolve?
Regulation will inevitably become more prescriptive. Supervisory authorities across Europe are already signalling that AI-enabled decisions, especially in credit, fraud management and customer communications, must be fully auditable and supported by high-quality data governance. Requirements for monitoring, model validation and bias mitigation will intensify as AI adoption accelerates.
Consumers, meanwhile, are becoming more aware of how their data is used and increasingly expect personalised services that do not compromise privacy. This means banks will have to strike a careful balance: offering more intuitive, tailored interactions while ensuring transparency, consent management and the right for customers to understand how AI driven bank strategies affect them.
In this sense, trust will become one of the most important competitive differentiators in the AI-enabled banking landscape.
Looking ahead, what AI trends will shape banking most in 2026?
By 2026, AI will be less of an experimental tool and more of an operational backbone for leading financial institutions. Responsible and explainable AI will become a regulatory expectation rather than a best practice, influencing how banks design and deploy new models.
Conversational and generative technologies will mature, supporting faster document processing, richer customer insights and new forms of advisory customer support, especially in digital channels.
Operationally, banks will continue moving toward more autonomous processes, particularly in areas like cash optimisation, service monitoring and fraud detection, where predictive analytics already show measurable impact.
Finally, omnichannel orchestration will evolve. Instead of channels operating in parallel, AI-enhanced analytics will help banks deliver experiences that are continuous, contextual and fully integrated across mobile, branch and self-service environments.
