We live in a time of digital gold rush, and the financial sector has thrown itself into the mines as one of the first. The scale of investment is breathtaking. According to IDC forecasts, European spending on artificial intelligence (across all sectors) is set to reach an astronomical $145bn by 2028. Banks don’t just want to go with the flow – they want to control that flow.
However, just beneath the surface of these impressive numbers lies a deep, systemic paradox. A global survey by DXC AdvisoryX conducted in 2025 paints a picture that should give every board member food for thought: while 77% of business leaders consider AI to be an absolute priority at board level, up to 94% of organisations face serious problems with its real-world implementation.
So we have a classic disconnect between intention and execution. Almost everyone agrees that without AI they will go out of business, but few are able to turn algorithms into real business value. Why is this happening?
Anatomy of an implementation dead end
Where does the error lie? The answer to this question does not require an analysis of the source code, but a look at the management culture. The research mercilessly exposes three major cardinal sins of digital transformation:
- Pushing responsibility onto technocrats: As many as 73% of companies believe that AI implementation should be handled exclusively by technical teams. This is a cardinal mistake. The result is technologically brilliant projects that, however, completely fail to address real market needs.
- Technology looking for a problem: 65% of organisations admit that they invest in AI without knowing what problem they want to solve with it. It’s like buying the world’s most expensive jackhammer and only then looking for anything in the office to demolish with it.
- Low level of human maturity: Despite massive investment, only 37% of companies have reached a high level of AI maturity. The rest are still stuck in a phase of expensive, scattered experimentation.
“The biggest paradox we see in companies is that AI is a challenge for IT departments, not for boards of directors. Meanwhile, decisions about which business problem to solve first must be made by the business – not technology teams. Prioritisation and coordination are also extremely important. Without this, even the best algorithm implemented by the best team will not deliver value.” – says Halina Frańczak, Managing Director & Market Leader at DXC Technology Polska
Europe’s geopolitical dilemma: Capital asymmetry
The European banking sector faces a unique geopolitical challenge. It is about technological sovereignty. When we look at the map of global financial outlays, a gigantic disparity is visible to the naked eye. Private investment in AI on the Old Continent amounts to around EUR 6.1 billion per year. At the same time, in the USA the amount reaches EUR 44 billion.
This sevenfold difference means that European financial institutions are condemned to use solutions provided by global giants outside Europe. McKinsey estimates that in order to build a truly sovereign and independent AI, Europe would need to invest as much as €15-20bn a year by 2030. Current private capital simply cannot bridge this gap.
For the banking sector, the stakes are highest. Figures from the World Economic Forum show that global bank spending on AI will rise from US$35bn in 2023 to as much as US$97bn in 2027. The table below shows how tough Europe faces in this global arms race.
Where are banks realistically using AI?
Despite the barriers, the transformation is progressing and banks are pinpointing areas where algorithms provide an immediate advantage. According to the European Banking Authority (EBA), the most common implementations are in areas critical to the security and stability of the institution: fraud detection, customer identity verification, anti-money laundering and cyber security.
DXC Technology’ s data confirms this direction. The three fastest growing application areas for artificial intelligence are:
1. cybersecurity (50%): Banks are using AI as a shield against increasingly sophisticated cyber attacks.
2. ESG (47%): Automating the analysis of giant data sets for sustainability and non-financial reporting.
3. Risk management and compliance (43%): Aligning operations with a dense regulatory network.
Most of these advanced systems come from external vendors. This raises the risk of so-called vendor lock-in. Banks are seeking to manage this risk in a mature way, with 75% of firms actively seeking technology partners and 83% considering a managed services model. Interestingly, more than the technology itself or building models, financial institutions value in their partners the ability to train staff and guide the organisation smoothly through the change process.
A new dimension of supervision: AI Governance
The introduction of artificial intelligence into the bloodstream of financial institutions is forcing the evolution of supervisory boards. Autonomous systems are entering banks faster than internal supervisory procedures are being developed. Although 98% of companies claim to have AI Governance structures in place, only 47% have implemented a mature supervisory framework that takes into account the human-in-the-loop model (human as the ultimate fuse in decision-making).
Banks face tough, practical operational challenges. The EU’s AI Act mercilessly enforces full accountability of algorithm-supported decisions and auditability of real-time learning systems. The good news is that a growing number of banks are handling these challenges efficiently, building the foundations for informed and secure oversight of new technology.
The European path: Trust over the blind pursuit of ROI
While the US or Asian market often relies on absolute speed of deployment, Europe is building its own unique technological identity based on trust. DXC data shows that 41% of European organisations prioritise data security over return on investment (ROI). In comparison, for only 28%, it is quick profit that remains the absolute priority.
This cautious, responsible direction has recently received powerful regulatory support. The European Central Bank’s (ECB) recommendations of 27 May 2026 clearly defined the rules of the game. The regulator has raised safety standards and sent a clear message to the industry: the question is no longer “whether to implement AI”, but “how to do it safely and responsibly”. Crucially, the ECB has officially classified AI-assisted attacks as a systemic financial risk, imposing an absolute obligation on banks to report their exposure to such threats.
Polish leader sets the standard
Against this backdrop, the Polish financial sector appears to be the absolute frontrunner and a testing ground for modern technology for the entire European Union. We are the leader in digital banking, which is perfectly illustrated by the success of the BLIK system, which processed a staggering 2.9 billion transactions in 2025 alone.
Polish institutions are not afraid of innovation – as many as 63% of them are actively implementing generative AI. However, looming regulations and increasing pressure on automation are forcing us to enter a higher level of operational maturity. As he accurately summarises:
“In the banking sector, where customer trust is the foundation of operations, a responsible approach to AI is not a weakness – it’s a strategy. Poland is a digital banking leader in the EU, with BLIK processing 2.9 billion transactions in 2025 and 63% of Polish financial organisations actively implementing generative AI. At the same time, the impending AI Act requirements and the growing pressure for automation pose the question for the sector not “whether to implement AI”, but “how to do it responsibly and at scale.” – Halina Frańczak concludes.
Based on: DXC Technology
