 Hi everybody. Nice to see you again. So, right today, we're in here in this room. We're going to talk about something that FINRA have presented at one of our special interest groups, the Ritech Innovation Special Interest Group. This is something that FINNO has four topics such as this, to talk about the digitization of regulatory rules and helping industry participants meet their compliance requirements and collaborate together. So, as part of the SIG, I think maybe six months ago or so, the FINRA team came to us and said, look, we're working on something that we would like to showcase to the industry. And they came to the SIG and they presented their pilot and then about a month ago, they came back and they said, look, we have a version. We have a first version live and we'd like to show it to the community, which we did. And we've had several of those. And today, the FINRA team is here to present, again, what they've done and to collect feedback. It's really, really important that we get feedback and it's important for FINNO, so it's important for FINRA. It's important for the community to get input on what you guys think, whether this is important, whether this is going in the right direction. So, I'm going to hand it over to the FINRA team. They're going to tell you a little bit more about the tool and how to use it, why it's there, what the purpose is. But also think of questions and feedback and input that you would like to give back to this team. Hi, everyone. My name is Alex Katchaterian. I'm a Director in the Office of Financial Innovation at FINRA. I'm here with my colleagues, Afshin Adubaki in the General Counsel's Office and Adi Rastogi in Technology. So, as Jane said, thank you for that intro. We're going to talk about our machine-readable rulebook initiative. So, I'll be real quick in my intro remarks here before we dive into the demo. First, we're going to talk about the goals of this initiative. What are we trying to accomplish and what are the use cases? Second, we're going to talk about how we went about accomplishing the prototype that we've launched on October 20th, how we got here. We're going to do a demonstration, as I mentioned, and then lastly, we're going to close out with what's next. What is the way forward? What is the future of this project and what does it hold and what kind of comments we're seeking? So, on the project goals, what are we looking to do with this machine-readable rulebook initiative? So, it has three components to it. First is the creation of a regulatory taxonomy. So, that's essentially a hierarchical categorization of key business legal and regulatory terms that you can use to filter and search through the FINRA rulebook. It's sort of like an index that allows a way to facilitate search functions. Second is the delivery mechanism for that taxonomy is called the FINRA rulebook search tool. Or first, that's exactly what that cross-filtration system is. Essentially, it allows you to look at different topics across various categories and narrow down your search of potentially applicable FINRA rules. This taxonomy and the application of the taxonomy to the FINRA rulebook was done as a manual effort by FINRA attorneys. It was brought to life through the user interface and the API through our technology team, and that's the third component. Speaking of the API of the machine-readable rulebook initiative, and that's what truly makes it machine-readable, is that all of this content and this very rich content developed manually, again, that can be pulled in through the API to map to your internal policies and procedures. You can get updates on rule content changes, which Adi is going to talk about. So, how does this help and what are we looking to do with this? It can help in a manual search for someone on a firm, let's say, who is looking to send a question to their chief compliance officer, but can maybe narrow down that question or streamline the question by clicking these different fields and find the potentially applicable rules. You can manage regulatory change through the API by getting automated updates and regulatory mapping. And broadly speaking, another goal of this tool is we think that it can support the broader REC tech ecosystem, not only for firms that offer these types of services, a new way of doing compliance, but also we're working with our regulatory counterparts around the world. So, tie that into how we got here. So, we've currently done this for 40 FINRA rules. These rules represent about half the rulebook views or half the visits really to FINRA.org more broadly. We created a pilot in 2018 after launching an initial special notice on this topic. We received feedback to move forward. We demoed the pilot, tested it with a number of firms. We sent out surveys and ran essentially a cost-benefit analysis or an economic analysis of the benefits to firms. We've presented and discussed, as Jane mentioned, to FINOS. We've also presented to major securities, trade bodies. We've created a working group of compliance professionals at broker dealers to vet and check and apply on our taxonomy and the user interface. We partnered up with a prominent REC tech firm to check the global consistency of our taxonomy terms, and we had an internal process of attorneys creating these tags, applying them to the rules, and then subject matter experts in the general counsel's office reviewing them. And with that, I'll turn it to another general counsel colleague, Afshin Adabaki, to walk through a demo, and I will go to the website here to show it. Great, thank you. Can this show the website that we have? I don't know if it's... Okay, that worked now. Great, thank you, Alex. My name is Afshin Adabaki. I'm a special advisor and associate general counsel and our officer general counsel at FINRA. I'm also co-lead on this project with Jaime Werkey, who's in the audience, and some of you may have heard from last night. Speaking of last night, for those of you who attended that event, there was a lot of excitement, I felt, with everyone who presented in terms of what open source means to them, in terms of what REC tech, and there are a lot of acronyms right out there, what that means and what that evolution will look like. And so, you know, I went back to the hotel room and I go, what excites me? I've been a regulatory attorney for 20-plus years. Rulemaking, as some of you may know, is not the most exciting job. You know, there are a lot of layers of review and there's oversight, and ultimately you get an approval, and then you have to start interpreting that rule. But when I think of this project, I'm like, well, this is what excites me as an regulatory attorney. For me, regulatory compliance is what drives market stability, which goes along with investor protection, and together you have a cohesive market that the folks apply compliance appropriately to. You're going to have a nourished and viable market. If you don't, it can be disastrous because there's no confidence in the market, you know, things fall apart. You know, obviously, historically, there are many examples, including some recent ones. So we think that leveling the compliance playing field is a great starting point because you can be reactive and enforce your rules, and that's one way of making sure that folks are compliant with this objective, regulatory compliance. The other way is to be proactive. How can you be proactive? Well, you can educate folks on what rules mean. You don't need necessarily the subject matter expert or an attorney who's practiced for, you know, decades to be able to give the initial guidance, which is, hey, I'm a firm, I'm starting up, and I want to understand, you know, what this world is, the FINRA rule book, the rules set, what are the various requirements and what specific rules apply to my type of business, and that's what I mean by leveling the playing field. I think ultimately you probably have to go to council for the final points of things, especially contracts and things like that, but to understand the basic compliance requirement, we think this is the kind of tool that will get you there. Many larger firms have these types of tools in-house. They have the resources, they've built them, and there are third parties that perhaps provide this type of resource as well, but to my knowledge, it's never come from a U.S. regulator. I mean, I would expand that. I think, you know, there are probably just a couple, a handful of international regulators who've done something like this. The first step is a prototype, but we think that the future is bright in this space. So what we've done here is we've basically curated our rule book. We've only applied it to 40 rules. Alex mentioned it's a prototype, so, you know, they're very labor-intensive, and that's why we're at this stage, and we haven't done the whole rule book yet. And we curated it. We took every rule, and by the way, our rule book goes back to about 80 years. That's where the first set of rules came around. Folks that obviously wrote in a different style may be using different language, a little more legalese, and over time, you know, there's been a drive to make it less legalese, more user-friendly, so that folks can comply with these obligations. So what we did is we went to each rule, read line by line, and basically deconstructed the requirements into tags, terms, taxonomy, you know, we used interchangeable concepts for that. And then we indexed them in the format you see before you. We created the summary topics, which are high-level fundamental concepts. There are eight of them, you know, overlapping concepts. For example, security type. You know, that's like, you know, that level between equity and debt, just as a general matter, and we'll get into some details of what else we did with this. But what the summary topics allow you to do is just immediately be able to find a related concept. Now, one would say, well, why can't I just search your rule book or go to Google or some other browser, search browser and just find what I'm looking for. And my response would be, this allows you to learn, educate yourself because it has a browse functionality. You may not know what you're looking for until you find it. And that may take a while. This allows you to look for a variety of terms within this universe and figure, oh, yeah, that's what I'm looking for. I'm interested in this other thing. Maybe I can overlap these two things and see what result I get. It's really advanced search function in terms of regulatory requirements. Just to give you an example, we're going to pick firm type. We use internally. Firms are aware of it. It's public as well. This is a classification that we use. So generically, there are capital markets and investment banking firms, clearing and carrying, diversified retail, trading and execution. That's your general concept. But as I said in the beginning, if you're a startup firm and you are interested, you're just going to do capital markets and investment banking, a narrower set of rules apply, you would think. And out of the 40 rules, if you go to the last page, I think there are about 32 rules here that apply. So right there, you've windowed down the results, and that's a great starting point. Whereas before, you had to hire someone, pay them the consulting fees to figure out, okay, what should I do to start? This does that for you. We do have a disclaimer. You shouldn't rely on it as illegal advice and regulatory advice. You ultimately have to make your own decisions and read the rules to figure out whether they truly apply to your business model, but at least it gets you to that starting point. If you're a diversified firm, which is the global range of products, all 40 rules apply. It'll come up in a second. There it is, all 40 apply. So right there, we think that's a plus. Now, what's that additional benefit? I always use the example of my first day or my first week 20 years ago. My assignment was go and find all the finware rules that have a disclosure obligation in them. And I was like, oh, it's easy. Back then, there was no search, but I was like, I'll just open each rule and type in control F disclosure. I missed about 30, 40%. The reason is we typically, we may not use the term disclosure in that form to mean, we use different terms to mean disclosure without using the exact term disclosure. So what we've done here is under obligations and duties, you have to uncheck those. We just went ahead out of the 40 rules and we identified all of those. If you click on disclosure, that have a disclosure obligation. Whether or not they use the word disclosure. That, again, it's a, you know, I think, plus plus, right? For that junior, not even junior. I mean, for any compliance person, it does that task for you. And the sort of next functionality that goes with this is what we call cross-browsing or cross-filtering. If you now want to pick, let's say, customer type, again, this is the higher level bifurcation, right, pretty basic, institutional versus retail. You have your retail customer type and you want to understand KYC, AML obligations, and you want it in the context of the account opening process. So you do account opening, servicing, or termination, and you get the two rules and by their names you can tell by their titles that apply to both know your customer, which is the term of art, as well as AML anti-money laundering. Now, you open these rules and you can do a reverse lookup. You pick any of the terms that we tag, which appear there as Alex is showing, and if you click, I'm just going to pick a random term, qualifications and registration, and it will now look up every rule that has that tag associated with it. So you can do that type of lookup as well. The other element of this is these detail topics that we came up with. And we had to, there's a dilemma, right? As we started this project, we're like, you know, is it better to overtag and then decide later, hey, we went too far and then reduce the number to make it more usable or, you know, under tag. We opted for overtagging and this is the result. So just to give you an example, investment products. At the higher level summary topics, you notice we had bifurcated, as I said, equity and debt. Here we go deeper. Every rule that has a specific investment product identified in that rule, we've tagged with the corresponding term. Direct DPP is the term of art. These are vehicles that have flow-through tax consequences. You know, again, two rules here just by way of example. There's some duplication, equity securities, you know, identified twice, both in the summary and the detailed topics. Just another example, foreign securities. Alex, if you go down, you know, which rules apply to foreign securities? Specifically, that's right there under fixed income. You know, it's a generic concept. Foreign securities is not really identifying the specific type of security, but there you go. Those are the rules out of the 40 that have a foreign securities element within them. You still need to read the rule to understand the obligation and where the foreign securities term appears and how it relates to you, but we've done it. And the other thing we did just in the prior example is when we did account opening, servicing or termination, that's not a term that we use. We came up with that term because it's a procedural concept, right? There's an account opening or onboarding process. Our rules don't specifically use that in terms of identifying what the obligation is. That term, but we came up with the phrase and we're telling folks these rules just to open account opening, servicing or termination under ACC, under summary topics. We're saying these rules right here have an account opening processing, sorry, account opening, servicing or termination component within them. Suitability is a good example. It's the first page, the obligation to make sure that if you recommend the security to a customer, it's suitable for them based on your analysis. Well, in our view, that's clearly both an account opening right when you're onboard, as well as a servicing concept. So we tagged account opening, servicing, termination to this rule. The last piece of this is free text searching. So in addition to providing the browsing functionality, you can search for the tagged concepts by free text search. You put disclosures and quotes. It does pull up the express references and rules to disclosures, but if you, you know, throughout, you know, there you go, 2360, I think is your, no, sorry. 2241. 2241, yeah, so that's the first instance where disclosure is showing up also as a tag term. You know, we use disclosure as an example. There are other examples as well, but you can do free text search, which will pull up the tag terms in that search as well as the express terms. So, you know, we think this is the future. I think lawyers will still have job guarantee. They have to read the rules and figure out what it all means, but we think this is compliance education, which will help everyone. And again, we're going to bring back this, the beginning subject, you know, regulatory compliance ensures stability, and that's key to the growth of the financial markets and obviously making sure that investors are protected. Thank you. Afshin for that robust legal detail. So as Afshin noted, we've essentially deconstructed the rules here, and I'm going to have Adi come up and talk about the API. So the machine readable aspect of this is all of this content, all of this carefully curated content can be pulled in through an API and mapped to your internal policies and procedures, and Adi can walk through the technical details of that. Hi, I'm an Adi Rastogi, and I'm an associate product manager in the data dissemination team of FINRA. So the API aspect of this can be found on the API developer center. Just to give a little insight into what API developer center is, it is a FINRA strategic initiative to support automation goals of member firms as well as broader financial industry, and it makes sense to have the FINRA rulebook API available here as well. So you can come in here in the documentation and here you can see the FINRA rulebook API. So FINRA rulebook API provides FINRA rule content, both the current version of the rule as well as the historical version of the rule. Along with that, the taxonomy that I've seen in Alex from OGCNO or FIRA that developed, we took that taxonomy also made it available in the API as well. So the rulebook API has now evolved into a TAG FINRA rulebook API. So in the API rule content that you will receive, now there will be additional fields that will be available, summary topics and detail topics that you saw right here in the tool. So in the tool, you can see summary topics and detail topics. The same thing would be made available via the API as well. Just to give an insight as to what that content would look, what the API response would look like, this is the sample request of the API. Of course, here we only have the sample data available and you can see that this is the sample response for 2019 rule. You can see the end dates is null because 2019 doesn't have an end date right now. You can see the detail topics here as well as if you're scrolling down, you'll see the rule content, the start date of the rule. So this way this API can be used to ingest the rule content and map it to the internal systems, to your own internal systems. Another endpoint that we have provided here is the post call, which offers slightly more features. So for example, one of the features here is that with the other endpoint, the get endpoint, you can get all the data. With this, you can also run sort of like a filtering but through the API. So you can see here what it essentially, what it means here is that give me all the rules that pertain to customer type retail. Now, this is the Federal Rulebook API, but when we launched this Machine Readable Rulebook Initiative, we also said that we are working on another API which will allow you to get auto updates for any changes to rulebook content. And I'm pleased to announce that two days ago, we launched notification API and the first use case for notification API is Federal Rulebook Notification. So what essentially happens here is that if a user comes in and subscribes to Federal Rulebook Notification API, what they will receive is, let's say, we go in and say, okay, something changed in 12.10. So the people who have subscribed to our notification API, the system that have subscribed to that, they'll receive a notification saying something changed in 12.10, updated. If a new rule is created, it will say action created. And then this API, with the response received in this API, the systems can call the Federal Rule Content API and download the latest current data. I'll turn it over to Alex to go over the next steps and be forward. So thank you, gentlemen. So we launched the rulebook, as I believe I mentioned, on October 21st, 2022. And we accompanied the three elements, the taxonomy, the first search tool and the API, and we accompanied that with a special notice requesting comments. The special notice essentially kind of walked through in greater detail, of course, everything we've covered here, what the use cases are, what the journey is. And we set out a series of questions to the public, asking how they interact with the Federal Rulebook, how they find the first search tool and the API working for them, how it's benefiting them and what the use cases are, et cetera. And then substantive feedback, could the tags be amended, expanded, changed, et cetera. And most importantly, perhaps future development. So what's the next phase of this project going to look like? So we have the special notice here. Finner requests comments on its machine readable rulebook initiative. And you can scroll through here and see all the details. And then at the very end here, so this is our economic analysis I referenced earlier, very end here is all of these questions. So the comment period is open until December 20th. So we welcome, we encourage your feedback, your comments. We do have a booth just on the main floor downstairs. Please come by. We have a few of us sitting here waiting to answer your questions and engage further. We can also open it up now. I think we might have a couple minutes. If there's any questions, folks want to ask about the initiative or the tools that we've put out there? Yes, sir. We're at the booth. Thank you. So I'll try to break the question down as I understood it. So the question is, are we doing something to allow an integration of this content that we created with financial institutions? And do you see some sort of future that's going to maybe allow some sort of automated machine executable regulation? I'm going to try to boil it down that way. So I saw a head nod. Answer the first question. Yes, we have the API and the API is what allows the ingestion of our content. All of this carefully curated content into firms internal systems. They can map it to their internal policies and procedures and keep track of regulatory change management through the notification API. So what we've partnered with throughout the development process, those are two of the biggest use cases that they're seeing from financial services firms, so that's number one. The second question is, what do we envision this sort of machine executable? Any question on what the future of this project will turn into should be addressed in the special notice. We put out a prototype at this point. What this will turn into, what the industry needs, what the industry is looking for and where it will go at this time. What we've created is a taxonomy to facilitate the search and the API to ingest that content. Where this goes down the line, it's going to depend on a number of factors, including the appetite externally as well as internally. So I hope that addresses your question, sir. Oh, Elliot, please. Yes, Elliot, you got a microphone coming your way. Within FINRA, we have a series of different initiatives using machine learning. And a lot of these initiatives are to try to make it so that a lot of the documents that we receive are processed in a standard way. So this is a lot of the reason for putting this technology behind it so that there's a standard way of analyzing the documents in a standard way of critiquing. Whether there are certain rules that are satisfied or there are procedures that are standard within there. So it's not only that, but also looking at initiatives to look across the industry in terms of helping us better understand what the standards are. Again, this is all done with technology. So there are separate initiatives in the direction that you're saying is desirable. Does that answer your question? Okay. And shameless plug, if you have more ideas towards that, come to the Reg Innovation SIG which meets monthly and put those ideas out there in the public. The issues are open. We put out a notice every time the meeting meets. You can record your ideas and just like FINRA come and discuss them together with us, we're out of time, but I think we had one more question over here. Do you want to answer that? You can come here. So we have been discussing it with our fellow regulators. The one thing is, using the SEC as an example, probably a decade ago, beyond that, they created some form of taxonomy application for the EDGAR filings, for instance. So there is some basis to work on there, but what we've done here is the next step and I think the hope is there will be a future where there's agreement that the terms that we're using are standard. So when we say disclosure, our fellow regulators mean the same thing, etc. And then hopefully globally as well. But there's some things that we will probably not agree on because we use different terminology and that's just the way it's going to be. I just want to add one thing before I know it's beyond our time. You asked about interpretation. We have stayed away from interpreting rules. I'm going to use disclosure again. If a rule is unclear whether disclosure is required, we have not tagged that rule as a disclosure type rule because that's interpretive. So that's, you know, we've tried to keep it within the four corners of interpretive gray areas. And that's why I said you're still going to need the lawyers to parse these rules. Thanks. Well, thank you everybody. Thanks for your questions. Thanks for your engagement. Like I said, we have more of these discussions at the Red Innovation SIG on a monthly basis. Please, please, please comment to FINRA. Post your comments on their website because this is important work and digitization of rules is a long, long process. So the more FINRA and other institutions like this our community gets feedback and input and a lot of the input in the creation of this is going to come from this community too, which is why we're trying to keep it very, very open. Thank you for coming and thanks FINRA for presenting. Thank you. Thanks.