I recently attended Money 20/20 Europe, and one theme stood out clearly: artificial intelligence is no longer a future concept in financial services, it’s here now, and scaling fast.

The tone throughout the event was focused, pragmatic, and forward-looking. Discussions were rooted in real use cases, not theory. Whether in breakout sessions, keynotes, or hallway conversations, the message was consistent: AI is moving rapidly from experimentation to execution, and it’s already reshaping how the industry operates.

Given our ongoing work helping financial institutions adopt and scale new technologies, the conference was a valuable chance to see how others across the ecosystem are approaching AI. I was especially interested in how organisations are moving from pilot to production, tackling everything from governance and strategic alignment to embedding AI in day-to-day operations. It was a great moment to share insights, compare strategies, and connect with others navigating similar journeys.

AI in Finance: From Theory to Impact

AI dominated the agenda, but unlike previous years, the tone was refreshingly grounded. The discussion has matured, less focused on promises and more on execution. The standout message was that AI is no longer a future bet; it’s actively solving business problems today.

Organisations like Klarna and Mastercard presented tangible use cases already in production: automated fraud detection that continuously adapts to evolving threats, AI-driven risk scoring models, and document automation workflows that dramatically reduce manual effort. Perhaps most notably, several firms shared how generative AI assistants are now handling a majority of inbound customer service requests, with high satisfaction rates and cost savings.

However, the consensus across the industry was also measured. There’s a clear line between using AI to augment operations and using it to make autonomous, high-stakes decisions. In areas like lending, underwriting, and portfolio management, AI is still largely used as a decision-support tool. Full automation remains difficult due to regulatory complexity, the need for explainability, and the ethical considerations that come with opaque models.

AI’s Organisational Impact

One of the most compelling themes was not just what AI can do, but how organisations are reorganising themselves to support it. Scaling AI beyond proof of concept is proving to be the hardest challenge, not because of a lack of good models, but because of fragmented execution. More companies are now recognising the need to centralise key AI capabilities, from model governance to data infrastructure, to ensure consistency, accountability, and scalability.

Governance emerged as a central topic. Institutions are increasingly navigating the tension between giving teams room to experiment and enforcing firm-wide standards on data privacy, model risk, and ethical safeguards. This is driving a shift toward centralised AI governance functions that can set direction, enforce controls, and ensure compliance, especially as regulations evolve and customers demand greater transparency.

Training and change management are also rising in priority. Leaders across compliance, risk, operations, and product are being brought into AI initiatives earlier, with many firms investing in organisation-wide upskilling. AI literacy is no longer confined to technical teams, it’s becoming a foundational capability across the enterprise.

Another notable shift is in how the CFO–CTO dynamic is evolving. Tech leaders are pushing to accelerate delivery, highlighting the cost of inaction. Finance leaders, meanwhile, are advocating for clear governance, cost control, and alignment to business value. The firms making the most meaningful progress are those where this partnership is tight, where innovation and risk management are being co-owned, and where strategic AI efforts are not just distributed, but coordinated and centralised around impact.

Navigating the Regulatory Environment

One of the most pressing undercurrents was the evolving regulatory landscape surrounding AI in financial services. The EU AI Act is a major step toward structured compliance, classifying many core use cases, like credit scoring and fraud detection, as “high-risk” and subjecting them to enhanced requirements around transparency, governance, and human oversight.

But the AI Act is just one piece of the puzzle. Financial institutions must also navigate existing regulatory expectations from the European Banking Authority, the Digital Operational Resilience Act (DORA), GDPR, and national supervisors. Several speakers emphasised the rising importance of model governance frameworks, with increasing focus on explainability, bias mitigation, and auditability.

As a result, leading institutions are embedding compliance into the design phase, not treating it as an afterthought. This includes robust documentation, clear accountability between data, compliance, and business teams, and treating regulatory alignment as a competitive advantage. The message from the event was clear: firms that build with regulation in mind will move faster, not slower. In a trust-driven industry, anticipating compliance requirements early may be one of the most strategic decisions a financial institution can make.

What’s Next for AI in Financial Services?

AI is unlocking major short-term opportunities for financial services institutions, especially in areas where efficiency and customer experience are key. Supporting service teams is one of the fastest ways to drive impact. By deploying AI-supported chatbots and assistants, institutions can offer 24/7 support for routine inquiries, reduce call center costs, and improve response times. These tools are increasingly capable of handling more complex interactions, freeing up human agents for higher-value tasks.

Another high-impact area is credit risk and underwriting. Traditional models often rely on limited data, but AI can incorporate alternative and behavioral data to assess risk more accurately. This enables faster, more inclusive credit decisions, especially for segments like gig workers or those with thin credit files, while also reducing default rates and manual processing time.

Fraud detection is also seeing major gains through AI. Machine learning models can analyze large volumes of transactions in real time, detecting anomalies and potential fraud with greater precision. This reduces both financial loss and customer friction by minimizing false positives and enabling quicker responses.

In the back office, intelligent document processing (IDP) is streamlining workflows like KYC, claims handling, and loan origination. AI can extract and validate data from documents quickly and accurately, reducing manual work, speeding up processing times, and improving compliance.

Finally, AI is driving more effective personalized marketing and cross-selling. By analyzing customer behavior, spending patterns, and financial profiles, AI can recommend relevant products in real time across digital channels. This helps financial institutions increase engagement, improve customer satisfaction, and grow wallet share more efficiently.

Final Thoughts

Money 20/20 Europe 2025 wasn’t about what might happen, it was about how to make it happen, and fast. AI, embedded finance, and digital money are becoming central to the next generation of financial services, not add-ons.

But none of this will scale without serious organisational change. It’s not just about adopting new tools, it’s about rethinking how teams operate, how trust is built, and how technology aligns to business value.

The firms that succeed will be those that combine bold innovation with disciplined execution, modernising their foundations while staying focused on real-world outcomes.