Beyond Borders: A Transatlantic Conversation on the Future of AML Compliance

Financial crime compliance often gets reduced to a checkbox exercise—something firms do because they have to, not because they understand why it matters. That's a problem, because the "why" is what makes compliance effective.

At Passthrough, we've built our financial crime compliance services around a different philosophy: compliance should protect firms and their investors without creating unnecessary friction. But we also know we don't have a monopoly on good ideas. Some of the most valuable insights come from practitioners working day-to-day in the trenches of AML compliance across different regulatory environments.

That's why I reached out to Pietro Odorisio, a business development professional at Cube Finance SA in Switzerland. Pietro has become one of the most influential voices in financial crime prevention on LinkedIn, where he regularly shares insights on emerging AML/CTF trends, regulatory developments, and the practical challenges compliance professionals face globally. At Cube Finance, he focuses on the development and market positioning of financial software solutions for regulated institutions, with particular emphasis on RegTech and AI-driven automation.

Pietro's unique background—combining business ethics with hands-on compliance experience and expertise in AI-powered AML solutions—gives him a perspective that's rare in our field. He sees compliance not just as a regulatory obligation, but as the operational expression of ethical boundaries in the financial system. 

I wanted to have a conversation that went beyond the usual compliance talking points. What follows is an exchange about the real challenges our industry faces. Pietro's insights offer a grounded, practical view of where financial crime prevention is headed and what it takes to do this work well.

Q&A With Pietro Odorisio

Mark Mangion: Pietro, you have a fascinating background—combining political philosophy and business ethics with financial crime prevention. What drew you to AML compliance, and how does your academic experience shape your approach to combating money laundering?

Pietro Odorisio: It's actually a question I often ask myself as well. My interest in AML began almost by chance during one of my early professional experiences, when I was first exposed to this field in a very practical way.

What I quickly realized, however, is that AML fits remarkably well with my academic background. I studied political philosophy and business ethics, and while compliance is formally a regulatory function, it is also a space where ethical practices, responsibility, and organizational behavior are shaped in very concrete terms.

In that sense, anti-money laundering is not just about following rules—it is about defining the moral boundaries of the financial system in operational terms: what is acceptable, what is not, and what responsibilities intermediaries have in preventing the misuse of their infrastructure. It is this intersection between regulation, ethics, and daily practice that ultimately drew me to AML and continues to motivate my work in this field.

Mark: You've worked extensively with AI and machine learning platforms for AML screening and monitoring. From your experience, what's the most significant way these technologies have changed the game? And what can they still not do that requires human judgment?

Pietro: The most significant change is the ability to analyze and correlate vast amounts of data in a very short time, improving the quality and timeliness of AML analysis. Thanks to machine learning, these tools learn from the day-to-day work of professionals and become progressively more sophisticated in their assessments. However, for this very reason, it remains essential that experts continue to guide and supervise them. Human judgment is still indispensable to set up and adjust the models, interpret the results, carry out qualitative investigations, and steer the tools in line with constantly evolving regulations and requirements—also because, ultimately, responsibility will always remain with people.

Mark: Let's talk about false positives—the bane of every compliance team's existence. How has AI improved this challenge, and what realistic expectations should firms have about technology's ability to reduce alert fatigue? Depending on the jurisdiction, how much human involvement do you recommend to oversee the AI mitigation of alerts?

Pietro: As I mentioned earlier, artificial intelligence should be seen as an evolution of existing tools rather than a complete replacement. Improvements in the refinement of alerts and the reduction of false positives can often be observed quite quickly, thanks to much faster and more accurate computational capabilities. However, the real progress comes over time: by training the system with internal policies and consistently flagging anomalies, the machine becomes increasingly precise, potentially leading to a substantial reduction in irrelevant alerts.

That said, this process cannot operate autonomously—it must be continuously monitored to prevent errors or biases in the models. Human oversight is therefore crucial, particularly at this stage of technological development. Moreover, since ultimate responsibility always rests with people, concentrating AI supervision on only a few individuals could be risky. A broader and more shared governance model within the compliance team is preferable.

Mark: You've advocated for building awareness that AML is not just a regulatory burden but a tool to protect individuals and the economic system. How do you recommend compliance officers communicate this to skeptical investors who see it as friction in the investment process? What examples of reputational risk have you identified that may be helpful for professionals in the industry to be aware of?

Pietro: Compliance should not be seen as an obstacle; on the contrary, it should be communicated in an increasingly simple and accessible way, through a sort of compliance design similar to legal design. If citizens clearly understood what compliance actually does within financial institutions, they would feel more reassured—and this would also be positive for the reputation of these institutions.

As for reputational risks, there are many examples of banks and intermediaries that have suffered serious damage due to weaknesses in their AML controls, exposure to high-risk clients, or the use of opaque structures. These cases show that reputational harm goes far beyond financial penalties: it can erode trust from clients and counterparties and negatively impact business in the long term.

Mark: Traditional sectors like banking have sophisticated AML programs and file most suspicious activity reports (SARs). But you've noted there are newer areas more resistant to preventive measures. Which emerging sectors concern you most right now, and what makes them challenging to regulate? Do you think the recent FAQs from FinCEN in the U.S. will prompt investment advisers to file more SARs?

Pietro: The crypto sector, which was born as a decentralized tool, is certainly one of the most complex areas from an AML perspective. I wouldn't necessarily describe it as "resistant" to preventive measures, but rather as still immature—it is a fast-evolving ecosystem, and we are seeing regulatory and technological progress almost every day.

Regarding FinCEN's FAQs, I don't believe their purpose is simply to increase the number of SARs, but rather to provide clearer and more precise guidance to market participants. In any case, the key point remains that—as in all areas of compliance—what matters is not the quantity of reports, but their quality and actual relevance.

Mark: Politically exposed persons (PEPs) continue to be a significant risk category. You've highlighted cases where PEP screening failed spectacularly. What are the most common gaps in PEP due diligence, and what should firms be doing differently?

Pietro: Mapping and identifying PEPs remains a genuinely challenging task. If you look at the FATF definition and then consider how it is implemented across different local legislations, it is easy to understand the complexity—particularly when it comes to certain geographic regions and the extension of checks to relatives and close associates.

First-level screening should in principle help prevent firms from missing these individuals, but achieving truly comprehensive coverage for all PEP categories is difficult in practice. One possible way to address this challenge is for firms to rely on specialized providers with deep expertise in specific regions, rather than trying to cover everything with a single, generic solution.

Mark: De-risking strategies—where financial institutions terminate relationships with high-risk categories—have become more common. What are the unintended consequences of overly aggressive de-risking, and how can firms manage risk intelligently without unfairly excluding legitimate customers?

Pietro: De-risking is certainly a problem, but we also need to be clear about what we actually mean by "high risk." Terminating a relationship should only happen for genuinely serious reasons, not out of fear or excessive caution. Regulators should take this into account and intervene in a more targeted way, so as not to unintentionally encourage blanket de-risking practices.