Artificial intelligence is having an increasingly strong impact on how banks manage risk, efficiency, and customer relationships. According to the latest report by consulting firm Deloitte, “2026 Banking & Capital Markets Outlook”, the banking sector will enter 2026 in relatively good shape, but under growing transformational pressure. Financial institutions will increasingly deploy AI-based solutions across the entire organization, while at the same time investing in more advanced systems to combat financial crime, whose scale and complexity continue to rise. The coming year will therefore be a time for decisions rather than further pilots – the advantage will go to those institutions that move from experimentation to full-scale implementations.
Between Optimism and Caution
While the U.S. banking sector will begin 2026 amid high macroeconomic uncertainty, mixed consumer sentiment, and persistent inflationary pressure, European banks are showing clear signs of recovery. According to CFRA Research data, by August 2025 their stock prices had risen by an average of 45% year-on-year, reflecting solid financial fundamentals and growing investor confidence.
At the same time, experts note that banks in our region can expect a revival in lending activity, driven by falling interest rates and a stable contribution from non-interest income. Although trade tariffs may lead to slight deterioration, the overall situation should remain broadly under control. After years of stagnation, institutions on the Old Continent can look ahead to improved growth prospects in the coming years, both organically and through consolidation.
Change is also on the horizon with the growing role of digital forms of money. In the United States, the debate is being driven by the development of stablecoins and their potential impact on payment models – especially after the adoption of the GENIUS Act, which opened the door for traditional institutions to engage in tokenized digital assets. In Europe, the direction will be shaped by the implementation of MiCA regulations, which introduce uniform rules for the issuance and trading of digital assets and increase transparency for service providers operating in this space.
“Banks in Europe, including Poland, have learned to operate under prolonged uncertainty – today their biggest strength is stability, while the key challenge is the pace of change. Falling interest rates create room to rebuild lending, but at the same time increase pressure for technology investments to deliver measurable results. In the background, we are seeing rising operational risks, including cyberattacks powered by AI tools, which further raise the bar for security and infrastructure modernization. After years of building financial resilience, the sector is entering a phase in which competitive advantage will be determined not by declared ambition, but by the quality of execution – including the ability to use AI at scale. We are seeing more projects move beyond the testing phase, although the speed and maturity of these solutions still vary significantly. This shift – from caution to development – will define European banking in the coming years,” says Przemysław Szczygielski, Partner and Financial Services Industry Leader for Poland, the Baltic States and Ukraine, and Leader of the Financial Institutions Risk and Regulatory team at Deloitte.
Rising Investment, Delayed Returns
The report’s authors point out that despite growing interest in artificial intelligence, many banks still run AI projects in a fragmented way, without a unified data architecture and clearly defined business objectives. In addition, Deloitte’s latest global “AI ROI” report shows that while 91% of surveyed organizations across sectors plan to increase AI spending over the next 12 months, returns on such projects typically appear only after two to four years – significantly slower than for many other technologies. One reason is that many AI initiatives are built on data that is not yet fully ready for such use, which complicates measurement of outcomes and slows down scaling.
“Financial institutions are increasingly aware of AI’s potential, but the biggest challenge is turning this awareness into solutions that work at enterprise scale. In many cases, weak data foundations, siloed systems, and a lack of clear governance models remain key constraints, leading to duplication of initiatives and making it difficult to assess impact. Our analysis also shows that returns on investment come gradually – the full effect of AI usually requires ordering data, processes, and ways of working. That’s why deployments that create lasting value need a combination of strategy, governance, and investment discipline. Only then does artificial intelligence stop being a series of pilots and become a key transformation tool,” emphasizes Tomasz Tarasiuk, Partner and Banking Sector Leader in Deloitte’s Consulting practice.
A key direction for future development will therefore be moving from isolated pilots to AI solutions integrated with the bank’s core operations. Progress in this area is often constrained by fragmented and incomplete data. According to the Deloitte Banking & Capital Markets Data and Analytics Market Survey 2024, more than 90% of banking professionals say that the data they need is often unavailable or too time-consuming to obtain. For 81% of respondents, data quality remains the main challenge. Experts stress that in the coming years, “agentic AI” will become increasingly important – systems capable not only of executing commands, but also of taking initiative, analyzing data, and carrying out tasks autonomously based on defined objectives and compliance rules.
Growing Risk Complexity
The more extensively AI is used in banking, the greater the need to strengthen security procedures and operational controls. The scale of the challenge is already evident: in 2024, U.S. financial institutions filed a record 2.6 million Suspicious Activity Reports (SARs), an average of 7,100 submissions per day. At the same time, the number of regulatory enforcement actions related to violations of the Bank Secrecy Act (BSA) and Anti-Money Laundering (AML) regulations is rising. These figures show that traditional monitoring and control models are no longer sufficient in the face of the growing scale, speed, and complexity of financial crime and the increasing regulatory pressure.
In the coming years, supervisors will expect banks to monitor financial flows even more effectively and respond faster to new forms of abuse – from money laundering in international trade to the use of digital assets and AI to create fake identities and obscure the origin of funds. As a result, institutions that fail to build more technologically advanced financial crime management systems may become increasingly vulnerable to financial losses and criminal attacks.
For this reason, sector players are increasingly using predictive analytics and machine learning models in AML processes. Tools leveraging generative AI (GenAI) are also being implemented to support KYC procedures. Integrated data architectures make it possible to combine transactional, behavioral, and contextual information into a single risk monitoring system, which shortens response times and reduces the number of false positives.
“We are seeing a clear shift toward machine learning–based tools in transaction analysis. These solutions can detect subtle changes in customer behavior and transaction flows, as well as identify suspicious activity patterns. GenAI is also being used more and more often to support the analysis of complex documents and accelerate many processes, enabling more effective use of existing human resources. This requires not only investment in technology, but also in skill development and process change,” notes Paweł Spławski, Partner in the Risk, Regulatory and Forensic team at Deloitte.
