This week’s COMPLY episode is final part of a discussion between John Zanzarella, PerformLine’s CRO, Ed Vincent, CEO of Lumio, Kunal Datta, Head of Product at Unit21, and Anna Fridman, Co-Founder of Spring Labs, as they take a deep dive into the transformative impact of AI on RegTech, exploring how AI-driven solutions are redefining compliance strategies for financial institutions.
This episode highlights:
- RegTechs are evolving to meet rising compliance expectations, offering faster and smarter AI-driven compliance solutions.
- Financial institutions are adopting proactive, data-driven strategies to strengthen compliance and stay ahead of risk.
- Emerging trends: how AI and automation are transforming scale by boosting efficiency and enabling smarter risk management
Show Notes:
- Listen to Part 1: https://performline.com/blog-post/episode-73-from-compliance-to-catalyst-ai-role-in-regtech-part-1/
- Listen to Part 2: https://performline.com/blog-post/episode-74-from-compliance-to-catalyst-ai-role-in-regtech-part-2/
- Connect with John Zanzarella: https://www.linkedin.com/in/johnzanzarella/
- Subscribe to PerformLine to stay connected to resources and updates: https://lp.performline.com/subscribe-to-performline
Subscribe to COMPLY: The Marketing Compliance Podcast
About COMPLY: The Marketing Compliance Podcast
The state of marketing compliance and regulation is evolving faster than ever. On the COMPLY Podcast, we sit down with the biggest names in marketing, compliance, regulations, and innovation as they share their playbooks to help you take your compliance practice to the next level.
Episode Transcript:
Jessica:
Hey there COMPLY podcast listeners and welcome to this week’s COMPLY Podcast episode as we continue the discussion between John Zanzarella, PerformLine’s CRO, Ed Vincent, CEO of Lumio, Kunal Datta, Head of Product at Unit21 and Anna Fridman, Co-Founder of Spring Labs, as they take a deep dive into the transformative impact of AI on RegTech, exploring how AI-driven solutions are redefining compliance strategies for financial institutions.
Ed:
John, you went through some examples there before of looking at different marketing use cases. Everyone approaches their marketing materials a little bit differently, I’m sure, right? They’re positioning themselves a little bit differently. They engage with partners in a different way. So I think that that probably applies equally there in the marketing space as well, that they need to be flexible and configurable and apply it to the unique use cases of a financial institution.
John:
Yeah, I think you’re right on, Ed. And we always say that same thing, that each of our clients has a different risk tolerance and risk profile. And, you know, we try to provide the technical solution and let them customize it to that risk profile and get the insights that they want and have the technology work the way they want.
But I thought what Kunal said was really interesting about, you know, the client doing their own level of validation on it, as well. And that’s one of the things that’s been interesting for us to see is that some of the banks are learning pretty much on the go. They’re building out model risk reviews. And I would encourage the RegTech vendors, we had years and years of going through third-party risk review to even work with a bank. And we built out processes and answers to questions and all this information and data that we would give them in advance. You really have to do the same thing with your AI program, with your models and be prepared to go through that diligence.
And in some cases, you know, banks haven’t done a lot of that before. So you really need to partner with them to help them understand what is the best way to validate and review. What data are you using? What data, if any, is going into an LLM?
So that’s been something we’ve been learning alongside our bank partners, but something that we’re also very proactive about. We want to have those conversations. Like I said, we have a CTO who’s very willing to jump on, like Kunal, and speak to a client and explain some of what’s going on. And that seems to help quite a bit.
Ed:
Yeah, Anna, you, similar to what John was laying out there, right? You talked about the proper application of AI and you spend quite a bit of time here in this space of customer interaction where automation and monitoring solutions certainly can yield, I imagine, efficiency. But to go back to the efficacy comment there that Kunal made, as well. Can you talk a bit about a couple of practical applications there within the space that you plan?
Anna:
Yeah. So I would just say, one of the things that Kunal mentioned is this idea of what can be done with AI. One of the small things that I can describe that’s useful for us, in our complaints analysis, is emotion. Being able to detect emotion and figure out how to, what does that mean? It matters in a customer phone call, for example, it matters what the customer emotion is. But so we have, this is an example, a very specific thing that we’re now able to do that is valuable, that was challenging before on a large scale. So figuring out these unique specialized niches for which AI is very good.
And to be able to properly tag your complaints, we constantly measure our output compared to human output, and we test it on the back end. So the ability to properly tag a complaint, with what regulation do they belong to? What is the entire substance of the complaint?
We’ve proven, frankly, through automation, we’re able to do it more accurately than a human. And having that to be a layer on top of your human interaction, a lot of regulatory oversight currently talks about the fact that you still need a human in the loop, building on those efficiencies, and we’re looking at the fact that when you think about complaints responses. When you look at from agent to agent, there’s variance from agent to agent, there’s variance to how they respond in the morning versus in the evening. And standardizing their approach to complaints, helping the background with automation to standardize that approach, to categorize consistently every day that, it doesn’t matter what time of day, it’s got categorized the same way.
The fact that you can aggregate from different modes of contact, right? From chat, from phone calls, from emails, and organizing these quicker. And creating a scaffolding response for your customer service folks.
So they’re able to do less of that day-to-day routine work and more focused on things like root cause analysis. Like that’s the next expected thing that’s expected from you by regulators is to figure out: why are you having the issues? Where are you having the issues?
So to free up those customer service representatives to focus on things that matter more, to take it to the next steps, to simplify the simple steps and focus on the things that matter.
Ed:
I love the themes there, right, that we probably all, I would say intuitively make sense and probably would get to quickly ourselves, consistency, accuracy. I love the morning versus afternoon or one agent to another, right? And you can actually create some, you know, some normalcy and some consistency.
But the idea of entering new domains like emotion, which we probably would all say that that doesn’t immediately come to mind when we think of data and analytics, right?
It’s kinda the 180 degree polar opposite, other end of the spectrum, whatever you want to say. But to think that we can apply some technology to that and interpret that is immensely powerful.
Anna:
Oh yea, if the customer is super angry, you want to know, right?
Ed:
Absolutely. I’m going to be a lot more careful now the next time I call in because I think that I’ve now got a profile out there, now it’s detectable by technology.
Well, so I appreciate that, and I like the last bit there of really giving people the time then to spend on the root cause analysis, right?
The understanding, the interpretation. That ultimately, that’s the value add. The value add is not the data munging, per se. It is the interpretation. And then what do you do from there?
So let’s get to that, this concept of being a bit more forward looking here, an emerging theme. let’s wrap together the practical applications, the strategic changes in how we think about compliance.
And let’s pivot beyond what we’re seeing today.
So Anna, I asked you there a bit about use cases that are in the mainstream, if you will. What are we doing today? But I’m going to challenge you and then John and Kunal to look even farther forward. What are the trends? What are the use cases that are likely to emerge that are not right in front of us, we’re not tackling at this moment, but what people need to be thinking about what’s coming soon?
Anna:
So I think I’m just going to call out a general category. I’m not sure if it’s not being attacked, but there was, we found this McKinsey report that talked about that 61% of post-contact work can be automated, but only 10-15% currently is.
So if you think about it, that’s a huge discrepancy, right? I think they estimated that’s a $6 billion worth of value, right? So this idea of post-initial contact automation for long tail workflows and organizing and structuring that data, I think is a deep opportunity for the future.
And there are a lot of things that would fall into that, but that is a huge space for potential for automation that is now available.
Ed:
That’s a staggering percentage and actual amount of spend or value that could be realized there. John, thoughts on your side around emerging trends?
John:
Yeah, so I would say one of the ones that we follow being in more of the sales and marketing compliance space is just consumer behavior. And consumer behavior is changing, right?
Similar to how we saw when the Internet became more widespread. We’re seeing a lot of people move from traditional search to large language model queries. I know personally I don’t use Google search nearly as much as I used to. I think if every one of us looked at our own company website traffic, the Google search would probably be down and you know, people are making financial decisions today based on prompts that they’re putting into ChatGPT.
And so we think about a future where our clients get visibility into the results from those prompts and can have insights into brand compliance and regulatory compliance. We also think there may be a world where there starts to be a paid advertising function to some of those large language models, and if so, the big banks will be some of the first companies that are asked to advertise on those channels. And we want to make sure that we’re there to provide them the same insights that we do today across their customer acquisition channels.
Ed:
Those are certainly some forward-looking thoughts. We’re living this journey now in our shop of moving from SEO to GEO and making sure you’re ready for generative interpretation by search engines and LLMs as opposed to the traditional search engine optimizations of the past.
And we think about where are we going to place our bets and make our investment, and that’s certainly an area we’re pivoting towards.
The idea of paid advertising into LLMs is a fascinating and incredibly intriguing one that certainly carries technical, but then compliance questions around it, but then also the ethical and the disclosures and everything that comes around that really interesting space.
Kunal, thoughts from your end around emerging trends?
Kunal:
Yeah, I think one thing that we’re seeing is scams. You know, I mean, AI on one side, it’s like being, we’re talking about this here in terms of like, like, how do we use this in terms of software, you know, but like fraudsters are doing the same thing, you know, it’s not like there’s one-off fraudsters here and there, there’s like full on organizations, right? And they’re using this technology in ways that they’re not inhibited by regulations, right? So they can move really, really quickly.
I mean, one thing I’ll say, like, practically today, right? If you’re listening to this and your organization still has, like, voice recognition, that’s done. Like it game over already, like that’s dead. You know, so like practically like, please like take care of that.
Other stuff though, like, I mean, voice agents for sure, right? Like agents actually taking actions on behalf of people. But on the scam side is impersonation scams, right? There was a recent report, I think it was the FTC that put it out where it was like, basically it’s the number of people who have lost people over the age of 60 who reported, this is reported to have lost more than $10,000 in a scam like fourfold since 2020.
So we thought COVID was bad, right? But like, and it was right with the rise of big butchering scams, know, long tail, like, two, three year relationships that these scammers will build. I don’t know if you ever get one of those texts that says like, Hey, have you picked up the laundry? or just innocuous like, Hey, how are you doing today? That’s what it is.
And that kind of scam, think, and now especially with video generation, being able to put, it’s absolutely insane. And I think if the fraudsters are using that technology without actually adopting that technology inside the institutions that control the flow of funds, the door has been left wide open in a city of thieves. So I think that, I don’t know if that’s a trend, you know, and something actionable, I hope folks listening consider that.
Ed:
At a minimum, there needs to be awareness to that. You’ve got to think about it. And I think how you respond to it and the action you take is probably a far more complex topic. But at least having awareness and calling attention to it, I think, is important.
I’ll ask folks here, Anna, I think it was you who mentioned earlier, the state versus federal dynamic and the pendulum swimming from one side to the other. Any thoughts from this group on the regulatory space there and trends around compliance?
Anna:
I can jump in. I would say, I think we haven’t touched yet on AI regulations. There’s been some commentary by federal agencies, but there have been laws that have been put out by states. And maybe that will change in the future, but so far, my read of the regulations is basically you cannot utilize AI, GenAI to circumvent existing regulations.
I think that’s been the undertone that I have seen in all of them is that you can’t use GenAI to circumvent fair lending. Just because you use technology, doesn’t mean that the laws don’t apply anymore. So make sure whatever you do with AI, GenAI complies with regulations. They’re still big on the idea of a human in the loop. Making sure that there is some oversight by a live human being, which makes sense. And obviously data security, there’s concerns about data securities and making sure that companies are following proper procedures, have proper policies in place. So I think that’s the sense that I’m seeing of all of the commentaries, they kind of come down to this.
Ed:
Any other thoughts on AI regulations? I’ve got one question that came from the audience. Before I get to that, any other comments on that topic?
Kunal:
Yeah, just one thing I think quickly on this is like, you know, regulations, you know, we can all speculate, but we have no idea what’s actually going to happen. There’s like tons of bills floating around across all kinds of AI regulation, right? And it’s changing all the time, right? We adjust as the technology is shifting and sure it’s lags. I think the one thing is like, consider like, what are the things that will not change?
We know the transparency, that’s not something that no one’s ever going to tell you ‘Don’t be more transparent’, like be less transparent, please. You know, that’s, that’s something for sure. I can take a bet on that and I know I’ll win. So, I mean, things like that, right. Or like ‘stop auditing your stuff’. No one’s ever going to tell you like, don’t keep an audit trail, right? Like for sure, keep an audit trail. No one’s ever going to tell you like the quality doesn’t matter. Like, of course the quality matters. Right?
So these things, like there are certain things that we can predict like with very high degree of certainty that will be true regardless of what the regulations are.
And I think as long as we’re sticking to those first principles, like it shouldn’t. Regulation or fear of regulation should not be an inhibitor to the use of technology. And the reason I say that, right? It’s not just like, yeah, everyone should use cool tech. It’s like, no, like bad actors are using tech, right? I talked about it before and clearly this is something that I care about, guess I’ve said it now, two questions.
But I do think it’s important to adopt the same technology that’s being used for the gains that it can provide, of course.
But also to make sure that we’re combating bad actors in that way. And then making sure you’re sticking with the first principles that you can be sure will be true regardless of any regulation changes that come.
Ed:
Yes, I was on a podcast earlier this week and someone asked me for a synopsis of how I would boil down a message to folks. And I said, move now was my message because to your point there, Kunal, right? The bad actors are moving now and they are investing massively in compliance and in beating compliance, and so maybe our audience isn’t going to get it perfectly right or they’re going to might not get all the way there, but I guarantee you that by doing nothing, you’re in a dangerous spot, because no one else is doing nothing. And so other people are moving. And so don’t let your perfect be the enemy of the good here, right?
Take an action, take a step, lean into your partners. And that gives you at least a fighting chance to defend yourself from these fraudsters.
Anna:
At a certain point, it becomes almost irresponsible not to innovate, right?
Ed:
Yes, look, and I think that’s it. Bring that back to the American FinTech Council, right? It’s about responsible innovation. And so the flip side of that coin is: you’re irresponsible if you’re not innovating.
I was at an event where the Senator from Ohio was talking about innovation and he was talking about the fact that it’s just table stakes, right? You have to bring that to the table in order to survive.
Anna, I’m going to direct our question that came in through the chat to you, and I’m happy for others to weigh in here. It’s around a fintech and how a fintech would get your sponsor bank and the regulators on board with these RegTech solutions. We talked a lot about tools and capabilities and the value of leaning into technology in the compliance area.
If you’re a fintech, how do you get your sponsor bank on side?
Anna:
So I spent a lot of time around sponsor banks and in that space and I feel like the good news is that they work so closely with their partners so closely with fintechs historically that they’re used to they’re constantly on the cutting edge, their partners are always pushing them to the next thing so the next thing so next thing so the good news is I actually think they’re quite comfortable with the process of coming to their regulators to talk about new things.
So I think in my experience, they actually move faster than large institutions because they’re used to the process.
And the regulators, what seems to me, are used to the fact that they’re bringing new things to the table. Obviously, they want to make sure that the vendors that they’re choosing, that there are processes that they’re going through, the vendor oversight is 100% tight, right? And that all the regular rules apply, but I think there is openness to innovation and they’re used to the process of bringing new things to the entities that oversee them.
Ed:
Yes, I like the terminology before that no one’s going to fight you for more transparency. So if you’re engaging with your regulator or your sponsor bank, you can be more transparent, you should have a sympathetic ear there with those audiences.
All right, we have used up all our allotted time here. I know we could probably keep going. So I appreciate the hearty discussion, the lively interaction, and all the thoughts that were shared by our panelists here.
We covered a lot of ground here from: having to respond faster in the market and being proactive to this trade-off between growth and expense and safety and speed. So I think that was a great jumping off point. And then really getting into the practical examples, whether it was the contextual analysis of videos or whether it was asking about evaluation sets of data, taking risk-based approach.
Hopefully our audience here heard some real specific ideas that they can lean into.
So with that, I’ll close us out here. Thank you.
Jessica:
Thanks for listening to this episode of the COMPLY Podcast.
As always for the latest content on all things marketing compliance you can head to performline.com/resources. And for the most up-to-date pieces of industry news, events, and content be sure to follow PerformLine on LinkedIn.
Thanks again for listening and we’ll see you next time!
Rhonda McGill