# Have No Fear? Fintech, Finance and the AI Future | DC Fintech Week 2025

- Channel: [DC Fintech Week](https://streameth.org/dc-fintech-week)
- Date: 2025-12-21
- Watch: https://streameth.org/watch/yt-ETYRWU4c424
- YouTube: https://www.youtube.com/watch?v=ETYRWU4c424

## Transcript

Well, first of all, I want to get us started by saying thank you. It's been so exciting to have so many of you as thought leaders in this space to be at our office here our headquarters too with AWS. We're really excited to host the event and I'm especially excited to talk with our great partner Greg Ruppert who's an EVP at FINRA who leads supervision and regulation and is really leading the way and how we rethink a lot of the supervision process and how we dig in to identify instances of fraud, but also how we just rethink efficiencies across our business. Um Greg, just as getting started, there's so much that you've done to really um set a vision for what the space would look like. I would love to just start the conversation by having you share a bit about what you see as the vision for supervision with the adoption of AI and certainly as we lead into agentic AI as well. >> Great. No, thanks. First of all, thank you for for having us. Uh thanks to all of you for coming. Um thanks for your partnership uh over the past several years. It's been great. Um I I think I'd start off by saying uh you know, what's foundation to our vision is that FINRA's not necessarily new to the space. Uh we were early adopters of the cloud uh with with with all of you. Uh and then uh you know, we've really kind of moved into using various types of models throughout the years on our surveillance platforms as we look at the markets, uh but even more recently uh with some of the more um shifts in data and how we can can ingest that from everything from customer complaints to the other other types of data that we get. As FINRA is a self-regulatory organization that we also uh are have the opportunity to have a closer relationship with the member firms that we regulate. And then it also kind of turns into a little bit of a partnership to provide them kind of the expertise that we have and and a feedback loop if you will of seeing what's happening out in kind of the the space that we're operating in this particular space with AI. Where are, you know, our firms using it? And if you look at we have some very large firms ranging to some smaller mom and pop. So, how do we kind of capture that information? So, we have a number of opportunities to be able to reach out to our firms and have discussions with them. Round tables, we have advisory groups, and then we combine that not only externally of what's happening but internally on using it ourselves. So, we're we're active users of various AI tools. And so, it's an ability to have examiners, investigators, surveillance experts that are actually using the tools that that we then talk to our firms about. So, really getting that that full circle perspective. >> Yeah, and it is such a different journey from where we started together, which was very focused on the technology side being infrastructure-related, optimizing storage, everything that you do with the consolidated audit trail as well. But this next advent is one that really comes to how the business is led and is really much more of a board level conversation. I'm curious what use cases do you see as most critical for FINRA or you most excited about in the next year? >> That's that's a almost a trick question. I think there's a there's a level of of use cases and I'll I'll point out thanks to a working group that we co-hosted together with with AWS, but we brought a number of the federal regulators, some state regulators, some other SROs in just to talk about use cases. And one of the things that we did in that event was we saw that there wasn't a standard nomenclature being used. So, we would often have and they probably like what you guys are having in your firms and your companies of debates of well, what is this technology or what is this tool and where is it as a use case? So, we started creating a, you know, at that point I think we had nine different use cases um talking to the other regulators, some international regulators again uh with with a workshop you hosted, uh but also with our member firms. We're up to about 14 use cases and we created an infographic and uh I published it on my LinkedIn account uh and, you know, again record number of views on that, but what we started seeing is firms were then able to use it from the perspective of, all right, well, we're comfortable using these two use cases and then let's like move forward in that area and then we have a more of a of a dialogue. So, really building that that common nomenclature. In that I think um you know, when you start to get into the compilation of data, we have we get a lot large amounts of unstructured data. When you think of the over 3,000 firms, uh we do over a thousands exams a year, we have our surveillance, we have investigations, um we get other data from government agencies as well as uh other vendors. So, how do how do we arm our staff with the, you know, the most amount of opportunities to spend their time either talking to firms and understanding their business or addressing the issues that we're identifying and if they're spending their time trying to to gather the data then compile it, that's not a a good use of time. So, I think from currently right now, I like those type of of tools um and then really diving into uh the models that can get sentiment like so sentiment analysis is where I think another area. So, ranging from, you know, we do a town hall, we get thousands of comments. How do you do some sentiment analysis and do some analysis there? Uh but then also we often ask the industry uh as well as the general public for feedback. Uh we have a new initiative called FINRA Forward. How can we be a more forward-leaning regulator? How can we modernize our rulebook? We've asked for comments and you can only imagine we got a lot of comments. So, how do you start to leverage those tools to get that sentiment analysis, uh and then bring in kind of common themes? And then we were able to turn that kind of over to our uh investor complaints that come in. Um so, huge volumes of those. And then more recently, and this is your pro tip to take back to your firm, um you know, the uh the complaints that come into the firms, there's a requirement that they would report the significant complaints alleging fraud or violations of rules to us. Um used to have to do random samplings of those. Now we can ingest all of them and do a sentiment analysis, and it spits out the ones that they they they kind of suggest that might be within our parameters. So, we're able to look at all of those complaints versus a random sample. So, it really has has allowed us to protect investors and safeguard markets even more. So, I think that's kind of will be as I see our our future state is is how do we get to the the crux of what what we need to look at to protect investors and safeguard markets faster, >> Yep. >> save firms time, and we save uh the kind of the the speed to market so we can get to the saving of investors faster. >> Yeah, and identify those needles in the haystack, but really have people focused on the issue as opposed to finding it. >> Yeah. >> Um you know, there's a lot that you do in that sense of collaboration across all of your member firms. Um what have you found to be some of the use cases that you think really help drive some of that collaboration together? So, you mentioned with the comments and feedback. Are there other examples of some of that industry collaboration that you're doing to to make sure that there's good awareness of where there might be points of friction that AI might help ease the burden on some of those firms, too? >> Yeah, so we've started a a number of round tables. So, we we start out by asking firms to kind of come to us and and where are they using it, where are they seeing it. Uh and our we do an annual report uh targeting to come out in in December uh where we kind of have the ability it's unlike most regulators that do our priorities letter ours our report actually takes the empirical knowledge we have through the thousands of exams and thousands of investigations of what we're seeing as potential risks or issues for firms throughout the year we combine that with our intelligence program and our ability to kind of interpret where we think that the issues might be coming and produce a report of kind of key areas AI was mentioned in last year's report we expect it to figure even more prominently where we're providing not only kind of risks issues but also what are effective uses and where we should be thinking about it so that's how we kind of start out the year but then throughout the year we're having discussions with our member firms on how they're using it any advice kind of that that we see from an effective practice and then also just leveraging it ourselves so we have kind of hands-on knowledge in that area lots of use cases coming in around the kind of the compilation data like so it's much like what we've been using but anytime that you're drafting reports and you know spending time doing that summary models like those are all very helpful and then you know what we would come back with where we see it as having a human in the loop to take a look at it and you know kind of make the assessment that the model's doing what's intended and you don't have the hallucinations but that's the kind of dialogue that that we've been able to have I think firms are starting to to think about it from a client-facing perspective and so we've recently hosted a couple roundtables with those firms asking what they're seeing what the issues are and so there's you know more to come as we would produce reports out like our annual report is what we're seeing >> Yeah. Anything surprising that you've learned through that process? >> Good such a great question. Um >> Sorry, >> threw that one out there. >> Yeah. Um No, I think I think uh firms are actually acting very responsibly in this space. Um where I think the by the bigger surprises that we're seeing is the like third-party Mhm. Uh what you know, some people call it vendor uh adoption. So, you you might have a vendor that's giving you a particular service. Nothing to do with AI, and then all of a sudden they've they've added an AI component on the back end. Um and then we've also started seeing fourth-party uh kind of risk identified. So, vendors have vendors, and then those vendors have an AI component. So, I think that's been some of the more surprising things I think for firms is diving into those contracts. That's not necessarily the technologies' >> Mhm. >> wheelhouse. You know, that a lot of those will get negotiated by a legal team. >> Mhm. >> Uh and so, really kind of making sure you know what what you're getting and then how you're using it. Some of them are pretty innovative, and I think as a user, people get excited, but then not thinking through uh the the subsequent risks about data protection. >> Yep. >> Uh and you know, what what what you're what service you're actually getting or what you're actually giving away uh by using that. So, that I think that's been some of the more eye-opening uh discussions we've had. >> Yeah, we're seeing the same thing. I think Gartner recently released a report, too, where they were saying by 2027, a third of all applications will have some agentic component to it. And 2027 is not so far away. So, I think we're seeing it evolve so quickly, where you said it's being adopted just organically within the DNA of companies. And it really is the same as dot com, and now we're here with AI. Um you know, it's so important to scale safely in the adoption of these technologies. Um and there's a lot that you've referenced in terms of the round tables that you're doing. There's a lot of experimentation and um different ways that you have to inspire innovation at FINRA. Um what did you see as the most important lesson learned or do you look at in terms of making sure that can scale safely? >> I think it's having the right people at the table and and building a a governance structure and I think preferably in some type of committee that you also you have technology, you have the business, you have your your general counsel's department, your lawyers, you have your privacy experts, you have your cybersecurity experts all together talking about kind of the what you want to do, what the acceptable use cases are and being able to kind of see eye to eye in that. This is not an area I think that you can offload and say this is just a technology problem. It's just that that partnership and making sure that you're thinking through all of the the kind of future risks that that it could be. How do you make it the safest as possible and then, you know, kind of where you are and so even before that I think as as we started to see kind of a lot of attention around AI, you know, all of your all of your employees are out there you know, consuming news and interested in it and making sure that they're not, you know, trying different services that are publicly available using corporate data on that. And so like you can't just kind of assume that your your, you know, you're in a a vacuum or you're in a silo but paying attention to what's out there and then making sure everybody within your entire organization is aware of what what the rules are, what the parameters are, what the acceptable use cases are and and that education is just goes so far but if you don't have the governance to process that I think that's what we learned from from the beginning is having that together >> Yep. >> is the key. >> Absolutely and that's one thing too just looking at how you thought about guardrails so that it takes out a bit of the human judgment too just in terms of that governance process automating that as much as possible so people feel safe in the environment without the risk of making an error there. You know, a big part of that, too, is also just the culture. And, you know, I've often, in the almost 8 years that I've had the pleasure of partnering with FINRA, said there's not a a culture that I've seen that's so driven and such a culture of builders and engineering. Now, that was when we were much more plugged in under the CIO's organization. This is a very different conversation. So, it's so important that we're bringing business leaders to the conversation to really just reshape what the business looks like. And one place that it can sometimes be difficult to do that is as you evolve culturally and also just upskilling so that people don't feel like that is a technology challenge. That's not for me. Perhaps, can you share just some of what you've done with FINRA to really bring along your leadership to be a part of the journey in in terms of really being able to move forward with a strategy that you have? >> Sure. I think it's it's a great point and one of my favorite early business books was culture eats strategy for lunch. >> Yep. >> So, you've got to have that that ability to bring the people you want with you. I think for us, we saw that it's toned from the top and one of the things we started out with is having it as a strategic enterprise goal. So, um, I I co-lead that goal with our CIO, Steve Randich. >> Yep. >> So, everyone in the organization can see the partnership between the business and technology. Uh, and then as a strategic goal, it's it's approved by the board, it's owned by the management committee and our C- CEO, Robert Cook. He's active in this space uh, and I, you know, I'll be on the lookout for a future blog where he's going to be highlighting uh, some of the how we're using AI, but um, he's he's active in these discussions. And then, um, you start to cascade that through through the organization. Um, we had a recent town hall earlier this year that set uh, not one, but two records. And the first record was it was the largest segment of any single town hall and that segment was on our use of generative AI. And the second record it broke, it was the most amount of people that came on to showcase our focus and use of AI. So, all aspects of the, you know, all all parts of the organization got to see other leaders, other users, technology, business, legal talk about generative AI and so that, you know, received a lot of attention and then you start to kind of move down through the organization with various aspects of training, various opportunities where we talk about AI use cases. So, we have, you know, weekly events where each business can get together and say this is how I'm using it and that that is huge. Like the the ability to say, you know, I'm using it for the same job that you're doing and this is how I'm doing it faster. I'm also a big proponent of that metrics will drive performance. So, I've asked our head of kind of AI and advanced analytics to start producing a report of how often people are using our internal chat. So, individual users each week by department. And then starting to look at other types of use cases that we can bring forward and what the top use cases are and then we have plans later this year for like an AI day where we're just going to, you know, build time into into everybody's schedule to focus on their AI usage just to really drive that attention and then they hear our leaders talking about using it. So, I think that's, you know, from a cultural shift we're really trying to get at and not just saying that you have to use it but arm people with the tools, the time, the training and and and educational opportunities and then just staying on top of it. That's the last thing I think I would say is this this technology is moving quickly as everybody knows. So, it's not something you can just build I can only imagine what the training plan looked like 2 years ago in this that has to be updated. So, we're we're super excited we're going to be offering an agentic training program that's going to be a two-part series within the next couple weeks. And so, we're we're trying to keep our workforce in line with all the great innovation happening in the space. >> That's so awesome. And also just that you really started with that education with your CEO and making sure that he was armed with the understanding to make strategic decisions. So, really just top to bottom the level of education that you've had I think has really made your workforce feel like they are a part of the journey and that they are empowered to use the technology in a safe way. So, that's really important. >> Yeah. >> Um and also just love you mentioned the AI days, you know, something that I'll also just share that I appreciate um that's a tradition at FINRA is your annual create-a-thons. They're not a hackathon, but rather what use case can we solve for and really build and take to production. And you know, doing that in hands-on format to reinforce a lot of that learning I think is so valuable. And I I know you've had some great outcomes from that, too. >> Yeah, so I I I brought my leadership team in for a day on kind of re- reassessing what our kind of strategic vision is and we had we had our our change management team come in and talk about change management and they actually provided us the the overall change management plan having used our internal chatbot um and then they said they wanted to test their plan against the accuracy. So, since it didn't include any proprietary data, they used a a level of ChatGPT to say pick apart this plan to see if it's if it's accurate. And so, you know, you're hearing everything from your your human resources and training teams uh leveraging the internal tools talking to a group of leaders and how they were using it. And then the person I think just she did an amazing job of she's like, "Okay, where did my prompt, you know, not go far enough? Anybody have any ideas? And three people kind of joined in and said, well, you could have asked this or you could have asked to double check here. Um, and so I think making it like a collaborative, you know, kind of effort that is hitting all aspects of the business and it's not just one particular part but everything we do could could benefit from leveraging some of these tools that we have. >> Yeah, agree. And I think the the collaborative approach is one that really helps you go faster as well. So, you know, as we're we're closing, I would love to just hear we have a broad spectrum from industry as well as public sector. What is your word of wisdom or call to action for this this audience? >> I >> [sighs] >> I it would have to be um, active being actively engaged in this space. Um, I'm an active reader. I I mean, I've already recommended one book but um, yeah, um, the the AI-driven leader was something was recommended to me from our technology group. Uh, and I I read it. Um, I've now bought it for my entire leadership team. Uh, other parts of our organization have it. I can see it permeating. Um, and that the book really was hitting on this this, you know, kind of this really isn't just a technology tool but it's a it's a shift in kind of your thinking and then back to your earlier part about shifting culture. Um, but but that um, book particularly was a call to action uh, for every leader to use it as part of their leadership tool. You know, how how much time do you spend, you know, doing your daily things of, you know, email triage or um, meetings or whatever taking up your time? How do you free up 80% of your time um, to devote to other things and how could you leverage AI to do that? So, uh, this kind of comes [snorts] with it, you know, they have its own uh, chatbot um, to kind of uh, be your chief of staff, be your uh, uh, coach, be your board of directors. Uh, and so I think that was that was helpful. I actually spent time with the author last week and it was like hands-on keyboard using the voice module and talking to it and so I think that's for me as as really the more I use it the more I see the opportunities and you can kind of take even your your your groups to like the next level. Um we you know we had and back to that sentiment analysis I just talked about you know the examiner was looking at it like how much time I saved and look how much more effective I was and I'm like what if we imagine a world where we create an opportunity for firms to use that as a service and so I'm thinking more strategically and as FINRA's move as a self-regulatory organization how do we arm our firms with more tools to reduce the cost of compliance and make them more effective and empower them to protect investors and safeguard markets. If I didn't understand the technology then I would not have been able to kind of come up with that that more strategic idea as the staff is is using it for more tactical reasons. So I think that active active engagement by all leaders is going to be the the the best way forward. >> Yeah, very wise and yes the only way to learn is really by doing and proving is by doing. Well, thank you for your active leadership and for your great partnership and excited for the future and what we can would drive together. Thank you. Thank you. >> [applause] [music]