Mihai Gaicu

Good morning, Mihai, and thank you for taking the time to speak with us about the upcoming Leaders in Finance AML event on 2 October. To start, could you briefly introduce yourself?

I’m Mihai Gaicu, based in Brussels. I’ve been with SAS for just over four years, after starting my career at ING Belgium. My background is in business engineering, with an advanced master’s in data science. I joined SAS after hearing the former ING CEO say “Instead of being a bank using data, we want to become a data company with a banking license.” That sparked my interest in data and led me to pursue further studies. At SAS, I focus on using AI and advanced analytics to fight financial crime. I currently work in a global role within our go-to-market and customer advisory team. We support regional teams on AML, sanctions screening, and the fight against fraud and financial crimes, while also shaping strategy, finding new trends in compliance, and liaising with R&D to keep our products aligned with the market.

At SAS, and earlier at ING Belgium, you touched on a striking remark from the ING CEO. You now work in AML, an area that has undergone major changes over the past five to ten years, with significant impact on the banking sector. In your view, what has been the most positive development in the AML ecosystem in recent years?

I first noticed this shift just before COVID. Once the pandemic hit, digitalisation accelerated and society became largely cashless. Even in Belgium, where we used to rely heavily on ATMs and traditional systems, everything started moving online. An example is the Itsme application. With just an app and your phone number, you can log into all government platforms or even open a bank account. The bank simply connects to Itsme, retrieves your data through an API call, and that’s it. Life became much easier.

From an AML perspective, cashless societies bring advantages. Digital transactions leave a clear footprint, unlike cash which can move freely and is harder to trace. That’s why some people, especially those involved in illicit activities, criticize cashless systems: they rely on cash to stay under the radar. Even today, you’ll still find many businesses preferring cash, sometimes to avoid taxes or reporting requirements.

Of course, this shift also pushes criminals to innovate in new, more complex ways, so the challenge doesn’t disappear. But overall, digitalisation gives banks, software providers, and law enforcement much stronger tools and visibility to fight financial crime than before.

Looking at recent AML developments, which have had the biggest impact on your day-to-day work?

The modern AML framework took shape in the early 90’s, with the PATRIOT ACT marking a global pivot in 2001. AML and CFT became much more cross-border, risk-based, and standardized. For many years, not much has changed for AML solutions. But in the past year, the biggest shift has clearly been the rise of AI, especially generative AI. Oracle, for example, announced a $300 billion investment with OpenAI, and its stock jumped overnight.

I’m not an investigator myself; I only have an indirect view of their daily reality through client stories. But as a software provider, we’ve seen AI reshape how we respond to customer expectations. Clients want smarter, more adaptive solutions that cut investigation time while also reducing false positives. This has forced us to rethink how to design our AML software.

For instance, we’ve moved from decades-old rule-based scenarios to advanced model-based approaches, such as post-alert scoring, to enhance detection and decision-making. This risk-based approach sets scores first, then determines whether something is suspicious. The algorithm learns from past investigations, including which alerts were true or false positives, and uses that data to improve over time. It can even assign probabilities, such as ‘this alert is 98% suspicious,’ helping investigators prioritize what to review and potentially report. And that’s just one example. That’s just one use case, and there are many others. Happy to show you more when we meet in two weeks.

If you look at the current developments, the quality of AI today is still relatively low compared to what we’ll see in the next 10 to 20 years. Would you say the same applies to the AML space?

The thing is, developments can change very fast. For example, in AI we already see use cases like real-time monitoring, where suspicious transactions are blocked instantly, or automated generation of suspicious activity reports based on all available information, which requires almost no manual effort.

The future is harder to predict. For years, nothing really changed, but now – driven by a more cashless and digital society and new technological tools, everything is moving quickly. Sanctions regimes are also becoming more complex, making it increasingly difficult to keep track of everything.

And in the regulated world, laws often lag behind. Even when the technology is there, it takes time for legislation to be written, and daily implementation remains a challenge. That gap between innovation and regulation is something we always have to keep in mind.

Looking ahead, what do you see as the biggest challenges in the coming years?

One of the main challenges I see with customers is cloud adoption. In Europe it has become easier, but in some countries, especially in the Middle East, public cloud is still a no-go. The concern is mainly compliance and data leaving the region. Private cloud is often accepted, but when data goes to a public cloud in Germany or the US, resistance remains.

As a company we continue to push for cloud adoption, because it’s the direction the whole industry is heading. More computing power is needed to manage today’s workloads, and cloud has become the standard. Just as virtual machines were the step forward 15 years ago, cloud is the next evolution.

Bridging this adoption gap will be critical in the coming years. Our solutions, like those of most competitors, are cloud-native. While institutions and cultures have often been reluctant to change, they usually adapt in the end, and we expect the same with cloud. Otherwise, you fall behind and drop out of the competition.

Looking at technological innovation, especially AI, how do you see the relationship evolving between AML professionals and the technical tools they use?

The adoption of AI in AML is crucial, especially in reducing false positives. Today, professionals spend a huge amount of time closing alerts that turn out to be harmless. If AI can take over that repetitive work, investigators can focus on complex cases that truly matter.

But as criminals also adapt and use data more intelligently, it becomes an ongoing race: build a wall, and they will build a bigger ladder. Criminal networks are becoming more sophisticated, which increases the value of human judgment. Investigators will need to adapt continuously.

Importantly, AI will not replace them. It’s not “man versus machine”, but “man with the machine”. These tools help us investigate better, gain sharper insights, and work more efficiently. In the past we relied on manual methods. Now we have AI to enhance, not replace, the investigator’s role.

In general you could say that people become more effective when using AI tools. Do you also think the work itself will become more interesting, since these tools give you many more capabilities?

Yes. If you look at the statistics, currently less than 1% of hidden money is actually discovered. With AI, I hope we can raise that to 1.5% or maybe even 2%. It’s not that AI will solve everything, but it can accelerate the process of identifying what’s right and what’s wrong. It will certainly improve things.

For my final question, a more personal one: how do you stay motivated in this field, given how resilient and innovative criminal networks are? You mentioned that only around 1% are ever discovered. Despite the efforts of tech companies, financial institutions, and the wider ecosystem, what keeps you motivated?

It’s a bit different from working with a software vendor, because you’re not directly involved in investigating crimes. That’s the role of law enforcement, who are out in the field every day. When I joined SAS, though, I realized something: it felt almost like the movie Minority Report with Tom Cruise. It had all these futuristic, 3D depictions of crime and fraud. And when I joined, they were showing me the solutions, with 3D tools to map networks in virtual reality. I was like, wow – it feels like being a modern-day Batman, fighting crime with these fancy tools and technology.

And I think that’s what makes it unique: you’re not just using technology for its own sake, but to actually make society better. Even if it’s only 1%, it’s still a meaningful impact. Sometimes when I talk to friends, many of whom went into consultancy, audit or investment banking after business school – I say, “Okay, I work in the AML field, we fight crime, we fight fraud.” And even if some of them don’t find it particularly exciting, for me it feels great to know I’m contributing to society, supporting technical communities, businesses, and individuals.

It’s really about having a tangible impact on society by contributing to it. Thanks, Mihai, for taking the time to speak with us today ahead of the AML event on 2 October. I’m looking forward to meeting you there.

Thank you very much.