Elliot Berman and John Byrne return with a global roundup of AML developments. They begin with reflections on the Wolfsberg Group’s latest guidance on suspicious activity monitoring, emphasizing innovation and model validation. The Basel Institute’s working paper on collective anti-corruption action highlights the importance of multi-stakeholder collaboration.
The episode also covers a UK report forecasting a surge in synthetic identity fraud, a Ponzi scheme involving fake AI investment tools, and a comprehensive look at the evolving stablecoin landscape. Canada’s 2025 national risk assessment is discussed, with insights into NPO risks, fraud trends, and the country’s role in global trade.
Additional topics include a local elder fraud case in Wisconsin, a Dutch bank’s challenge to an AI-related AML fine, and upcoming AML webinars and interviews. The hosts close with a tribute to workers ahead of Labor Day and a reminder of the importance of vigilance in financial crime prevention.
AI, Fraud, and Global Risk: AML Insights from Around the World - Transcript
Elliot Berman: Hi John. How are you today?
John Byrne: I'm good. Elliot celebrated my grandson's second birthday with a nice trip to the National Zoo, so that was fun. And since we are dealing with, uh, zoo related activities in Washington, I thought it was appropriate. Uh, a nice day that we had.
As we go to, as we go to record. Sadly, another shooting this time in Minneapolis. I will just say there's way too many of these and we, it's just happened in the past couple of hours, so, we'll just, just simply leave it there. So anyway, I just wanted to mention that.
And, um, I think we'll start with a couple of international issues. We've talked many times about our friends at the Wolfsberg Group. Back in 2024, they issued a report. On, um, how they believe an effective monitoring for suspicious activity should be handled. So they, they put that report out in July of last year. So this year they've updated it with the second the second report, if you will, and obviously take a look at it and read it closely.
Briefly, the group framework that they're focusing on they're talking a lot about innovation are three pillars that they mentioned, and they go into detail on transition and validation, model risk, balance with financial crime, risk, and explainability to demonstrate transparency and coverage and effectiveness.
So take, you know, take a look at the report when you have a chance. Obviously, as we all know, uh, Wolfsberg international group of financial institutions their guidance, their statements are not only relevant to our community and industry, but supervisors and regulators in, uh, various jurisdictions also find them very useful.
So I would, um, uh, I would just urge people to take a look at that.
Elliot Berman: One of the, uh, the pillars you mentioned is validation and, uh, as many people who listen to us know model validation has been around for a long time in various forms. About 15, 20 years ago, it made a transition from principally, uh, IT related things to, uh, many things that.
People didn't really think of as models, uh, ways to calculate interest and other things like that. AI is likely as it gets deployed in the compliance space and in the monitoring space, is likely to be determined by the regulators to be a model. And it'll be interesting to see what the validation process for large language models and things like that turns out to be. 'cause um, that's part of the being able to explain it is, uh, how does this thing really work? So be interesting to see where that goes.
John Byrne: Yeah, that makes sense. So we're gonna talk in a little bit about a report that touches on some of the challenges with AI, which is something obviously we're, we're grappling with. The Basel Institute on Governance, another group we've been able to talk to several times has put out a new, working paper. This is Working Paper 57. The title is Mapping and Strengthening the Evidence Base for Anti-Corruption Collective Action.
And the Report explores how you can analyze collective action, uh, as they say in a way that respects diversity, but still allows for consistency and comparability. They hoping that a collective action will bring together, diverse stakeholders to tackle corruption risk that no single actor can resolve. And so that becomes a goal. Um, it is, uh, they vary widely in form and goals and context that they admit, the three authors. And they say that makes it a challenge to draw meaningful comparisons, but I still think that the value proposition is there about the need for collective action as an anti-corruption tool. You know, sounds, sounds sort of obvious, but. It doesn't always occur, and I think that's that's important.
But anyway, take a look at that. On the Basel website as we said before, they do excellent work on a whole series of things related to corruption and asset recovery. They do the Basel AML Index, so it's really a good resource for AML professionals.
Elliot Berman: There was a lot of work reported in the working paper, about the effort to be able to measure the effectiveness of, uh, of collective action on an ongoing basis. And, uh. I think for people who are familiar with it or just learning about that process and that framework, it's a really good in depth look at the challenges of, uh, getting the right people together and measuring their ability to curtail corruption.
John Byrne: And as we mentioned with Elliot mentioned AI the Organized Crime and Corruption Reporting Project, another great investigative journalist organization that does excellent work in our space and others issued a report, a report or actually reporting on a report.
Uh, the Global Bank Fraud Risks are Set to E scalate. And so this new report, um came from the UK based consultancy, Juniper Research. And the study looks at fraud detection and prevention in the banking market. It projects 153% jump from an estimated $23 billion in 2025. And they say one of the biggest risks is synthetic identity fraud, which as you know, is when, uh, criminals make fake personas by taking stolen, real, and fabricated details. And they can pass standard checks because they include partially authentic data and that becomes hard to detect. But they also say that the rise of artificial intelligence is fueling the problem, enabling faster and more convincing fake profile.
Quoting from one of the analysts at Juniper, synthetic identities can stay under the radar for longer and drain more money from banks, unquote. Quoting again, institutions must rethink static verification and introduce continuous monitoring if they want to keep pace. So that report is available on the OCC RP website.
Elliot Berman: Earlier this month, we posted to our website a podcast about synthetic id. So for those in the audience who'd like to, uh, know more about it, uh, or see if their current knowledge is up to the minute, uh, they can find that on our website, under the, uh, podcast, AML conversations tab.
John Byrne: Now, um, are friends of the IRS announced a major action and, uh, and I know you wanted a report on that.
Elliot Berman: Yes. So, in the Southern District of New York Federal Court there was sentencing for someone who had pled guilty for running a Ponzi scheme that involved a an a hedge fund that was supposed to be using an a, uh, a sort of a mystical AI algorithm to improve returns on in people's investments.
Of course, there was no such algorithm and, uh the person was caught. IRS-CI was actively involved in the investigation along with the FBI and, uh, actually the SEC as well because the person was a registered investment advisor. Gonna serve 30 months has to pay about a little over $4.2 million in restitution. The total scope of the fraud was about 5.7 million.
John Byrne: Excellent work again by, um, by IRS. Also there was a report related to stablecoin that I think you, you identified.
Elliot Berman: The organization's BRN, they put out a, a report on stablecoins. The report talks about 2025 being a pivotal moment where stablecoins are moving from, uh, crypto trading to part of the critical infrastructure for global commerce. They cite the passage of the GENIUS Act and they look at nations, uh, worldwide that are developing frameworks of different types, but all related stablecoin.
So there's comprehensive oversight formats that the US, Singapore, Hong Kong. There's restrictive approaches that are favoring sovereign digital currencies like, uh, what China and India have developed. There's been a significant market transformation. The stablecoin market has, uh, scaled dramatically.
Trading volume as of May, 2025 was $625 billion. And projections are that by 2030 the range of transaction volume could be somewhere between, uh, $1.6 and $3.7 trillion. So we would see a noticeable part of worldwide transactions being in stablecoin.
Uh, they also talk in the report about what they describe as the global laboratory effect where a lot of different countries and organizations are working on it. They indicate that that Asia leads global stablecoin adoption with 56% of the regional institutions already operational, which is the highest rate around the country.
As you know, the GENIUS Act has stablecoin provisions and, uh, they also talk about the two major stablecoin currencies that have significant pieces of the marketplace, and they talk in some detail about the difference in their approaches.
Uh, that's at a level of detail beyond the scope of our, our chat. But for people who are interested, uh, recommend that you take a look at the report.
John Byrne: Excellent. Also we noticed, uh, that FINTRAC issued their, 2025 assessment of money laundering and terrorist financing risks in Canada. A lengthy report over 125 pages. Uh, I just wanna pull out one item because I just did an interview that we'll be publishing after Labor Day, posting after Labor Day. On NGOs and charities and then obviously want to hear what jumped out at you, but according to the, to the many, uh, areas covered by the report.
The vast majority of NPOs in Canada present little or to no risks for money laundering and terrorist financing with only a small subset undertaking activities that are vulnerable to terrorist financing abuse. These risks should be evaluated on a case by case basis based on credible information and taking into account mitigation measures.
NPOs are subject to administrative oversight at the federal, provincial, and territorial levels. Particularly regarding incorporation or tax status. Very similar to how charities are handled in the US and as I said, we, I had a excellent conversation with Ashleigh Subramanian-Montgomery from the Charity and Security Network.
And this is a constant theme for NPOs to explain to policymakers and to regulators and to financial institutions what. Are requirements, what are best practices so that not all charities are treated the same way. But anyway, obviously there's a lot more in there, Elliot, what, what other things in their, their risk overview struck you?
Elliot Berman: Uh, they called out, uh, the fact that because Canada has a strong, open and stable economy and financial sector, that it makes it an attractive source, destination and transit point for proceeds of crime. So the good news is financially Canada's doing fine. The bad news is doing it in an open way, which we all think is a good thing makes you also, more attractive as the pathway. Uh, another thing they pointed out is that there has been significant uptick in third party enablers really offering what we've called in other conversations, money laundering as a service, and that's something that they're taking a serious look at.
They indicated that for terrorist activity they think that uh, most of the financing landscape overall remains low volume and low value, and that most of the terrorist attacks in Canada over the past decade have been perpetrated by ideologically motivated loan actors and that their vulnerability remains relatively stable.
So, uh. I don't know that the United States could say the same thing. In their looking forward section talk about the, uh, that fraud remains a dynamic and growing threat. And you and I seem to be talking about that every week because we're seeing that from many different angles.
And they also talked about the fact that the threat is increasingly enabled by new technologies such as artificial intelligence. And so again that is a global theme. We're seeing it from, uh, many different angles and, uh, I don't think we're gonna see that slow down, uh, anytime soon.
John Byrne: That's right. And given Canada's ability to withstand the apparent trade war that's going on with us and them, I think it's gonna be interesting to see how they handle and they've mentioned some of, they don't mention that specifically, but they do mention import export companies. They have over 215,000 import and export companies of various sizes.
So there's gonna be a lot of issues. With customs issues, um, tariff issues, tariff evasion, all that sort of stuff. So it's the global trade trade that they're gonna enter into that they don't do right now will be with other countries or they'll step up their engagement with other countries given what we are apparently doing to them from a trade standpoint.
Elliot Berman: Yes. And I think that will be a problem globally. A new wrinkle being the variability and the range of, of tariffs and how that fuels some of the potential crime and other risks that you talked about. And it'll be interesting the next time the US publishes their national risk assessment, how they take that into account.
John Byrne: That's an understatement. Um, there will be a hearing on September 9th at the House Financial Services subcommittee. I don't have it in front of me which subcommittee, I apologize. Where Director Gacki from FinCEN will be testifying. All of us in the AML community should be paying attention to what, uh, members of the committee are asking FinCEN, just given all the things that have occurred since January. , It should be an interesting day, let's put it that way.
Elliot Berman: Agreed. So there are two other things that I just wanted to mention. One is, uh, I saw it in the Milwaukee Journal Sentinel a fellow was, uh, indicted by a grand jury here in the Eastern District of Wisconsin for an alleged money laundering scheme that targeted elderly victims and involved taking cash and gold from them. More than $500,000 worth of value was handed over by victims to people whom they believed work with the federal government.
So, uh, that indictment just was handed down the end of last week. Uh, we'll see what happens. Uh, the person did plead not guilty but we're at the front end of the legal process now that we've gotten past the investigative process. A lot of these people were older, so elder, elder financial abuse, but also, uh, impersonation of government, uh, officials, all different kinds of things. So, I guess everybody on the call knows, but continue to share with everybody, BEWARE.
And the other one is, uh, comes out of the EU. The Dutch Central Bank has issued a, uh, $2.6 million. I'm sorry, million Euro, which is about $3 million fine to an online banking platform called Bunq, BUNQ. I don't know that I would've picked that name for a financial services company, but it probably means something else in Dutch, for serious AML deficiencies related to transaction monitoring.
The central bank. Uh, indicated that, that, uh, Bunq was using an AI based process for transaction monitoring, and that, uh, while it was picking up on suspicious transactions, it wasn't consistent in flagging those to officials. And it wasn't clear also that they were that they were picking up on all the ones that one would expect.
Our friend, uh, Sarah Beth Felix in an opinion piece on AML Intelligence, uh, talked about this case and why is everybody talking about it? Because Bunq is not just agreeing to pay the fine and fix things, they're actually gonna challenge it. And this could be at least the beginning of a lot of conversations about how effective are aI based systems in this space, and how much can organizations rely on them to do the level of detection that's necessary to protect the financial system and to catch the bad guys.
But Sarah Beth commented that at least from reading the the report put out by the central bank in connection with the fine that one of two things was happening here. Either the AI model was poorly trained. It's one thing to have an AI model, it's another thing of what do you give it to learn and how do you, how do you teach it to make good decisions or the things that were coming outta the system and being flagged for analysts to look at were being looked at by ineffective analysts.
And so those are always two pain points, uh, whenever you use a system, whether it's a AI or something else.
John Byrne: Uh, by the time we post the August webinar will have been completed. It's on the growing, um, amount of fraud. We're gonna talk both domestically and globally. So obviously at some point after the live program, we'll be sending out recordings.
What are we doing in September?
Elliot Berman: September, we're doing a program on enhancing your KYC program. And, uh, I have, uh, a number of excellent experts lined up. So, uh, we'll be talking about, I'm sure many of the things that you and I talk about every week because, uh, KYC has a lot of different facets to it these days.
It becomes more and more important to know who your cus know who you're onboarding so that you can really pay attention to higher risk customers who you have in the house at a different level than, uh, folks who don't present as much of a risk. So we'll be doing that. That is September 25th and like most of our webinars, that'll live stream beginning at 1:00 PM Eastern Time, and it'll run for an hour.
John Byrne: Great. Um, we have a couple of, uh. Interviews that are coming up next few weeks, I, I have scheduled the interview, uh, with the Antiquities Coalition that that'll happen in early September. I just wanna give folks in the US Monday is Labor Day. I want to thank all the workers that do so much to make this country move and, uh, you know, we're with you.
And we are part of what you do too. Uh, I think we should recognize on Labor Day we can do barbecues, but we should also recognize the importance of people that work hard for a living.
Elliot Berman: I absolutely agree with you. It's, uh, it's easy to forget or not pay attention to all the things that have to happen for our daily life to go smoothly. And a lot of that has to do with a lot of folks who put in a lot of hard effort. ,
John Byrne: Stay safe, man.
Elliot Berman: Yeah, you too. I'll talk to you next week. Bye-bye.