Podcast: AI in finance: Massive innovation, massive disruption, massive opportunity

I’m excited to introduce the fourth episode of season six of our Financial Futures podcast. It covers one of my favorite topics – AI and innovation. Listen to it now to hear me to talk to Ed Abbo from C3AI about why AI could turn capital markets on its head, as well as:

• Cryptocurrencies and tokenization
• Blockchain
• ESG
• Cybersecurity and anti-money laundering

And that’s just a few of the hot topics that intersect with AI. It’s a fascinating discussion. So, click on the button below to start listening.

And explore more of our AI insights on our Innovate for Value page.

Miss the last episode? Frontrunners of modernization: BPaaS and the cloud

Be sure to listen to the next episode: Fighting fire with fire: Risk and compliance in a chaotic era

FULL TRANSCRIPT:

ERIN DNAGLER: It feels like it wasn't that long ago that AI was just a fiction. But after years of development, this once sci-fi concept has been integrated into numerous solutions across industries. And AI, isn't the only game changer that's shaking up business models. Robotic process automation, machine learning and cloud technology are also poised to revolutionize the way that financial institutions operate.

ED ABBO: Financial institutions in particular have never seen such a numerous number of technology vectors change on them at one point in time. So what that means is massive innovation is possible. And with that massive disruption.

ERIN DANGLER: There are even innovations in the assets themselves, which could potentially lead to a complete re-imagining of the way trading takes place.

TONY WARREN: It would turn the industry on its head. If it really took hold because it would become a replacement technology end to end. And so, everything would be done potentially very differently.

ERIN DANGLER: This is financial futures, the podcast that charts the frontiers of fintech innovation. In this series, we'll be exploring the opportunities and challenges facing the capital markets and diving into the trends that are reshaping the way institutions operate in this rapidly evolving industry. I'm your host, Erin Dangler. And today we're discussing AI and innovation with president and CTO of C3 AI Ed Abbo and executive vice president and head of strategy and solutions management at FIS, Tony Warren. We'll be exploring the challenges the capital markets are facing and finding out what some of the inefficiencies are in the current operating model. We'll ask which technologies are set to disrupt the capital markets for the better and discover the steps institutions need to take to get ready for a digitized future. And we'll find out how technologies such as AI and robotic process automation will benefit firms as well as their clients. To start us off, Tony Warren tells us about some of the internal and external problems that institutions in the capital markets are currently facing. Tony, what do you see are the challenges emerging in capital markets today? Can you set the scene for us a little bit?

TONY WARREN: Yeah, absolutely. So I think it's a pretty consistent message that we get back of servicing this industry. And firstly, it's a rapidly and fast changing environment that we're living in, uh, with technology, you know, right at the core of that, you know, when you think about change, I mean, the world has been turned on its head pretty much in the last two years. And so everybody needed to react very quickly with great agility and great speed in order to, um, keep the businesses going and maintain their services. So we see, you know, rapidly changing environment, uh, a need for speed and the need for agility as an absolute key thing. And, and with technology is the driver behind that, the second part of it, and really hand in hand with it is, our clients customers are becoming increasingly sophisticated with their demands. So it's no longer acceptable to be waiting for data and services for the sort of batch cycles of yesteryear, everything and the expectation is for online real time data. And the data to be in a format that's very easy to use to read and then to transmit, to drive other services. And so, you know, customers want really an almost smartphone like experience in everything they do. Uh, the third piece of it is then we are an extremely regulated environment. And again, you know, thinking about the pandemic who would ever have thought two years ago that our whole workforce across the world would have been pushed into, uh, working remotely. And so from a regulation point of view, I think it's that that has really grown. I think there's a whole new focus on, uh, cybersecurity to keep the bad guys out. And especially now we're in, in this more distributed environment. And then of course is surveillance. So everything from anti-money laundering to trade surveillance, to communication surveillance across everything that firms are doing. And then finally, of course, it's the focus on growth. So it's making sure that the balance is correct with the cost of doing business embracing technology. Um, so firms can move themselves into a position of sustainable growth and ongoing success.

ERIN DANGLER: Yeah, I think you've definitely covered it. That's a lot to be happening all at once. Fast changing environment, customer demands, increasing customers, getting more savvy, cybersecurity, all of that. So as financial institutions are rushing to keep up or to grow with this huge curve, what are you seeing as some of the most common inefficiencies that financial institutions in the capital markets want addressed?

TONY WARREN: I think it's coming from older siloed technology to the next generation. So I would put almost at the center here is this evolution of infrastructures. So in order to get to the next generation, the data has to be removed from siloed engines, which were successfully processing different departments within an institution. But now what people need to see is the ability to digitize the data end to end so from the front middle to back office and back to the middle to the front and anywhere in between, and it's being able to get to that state and then hand in hand with that is the distribution of it. So be able to move to more cloud-based environments, whether they are privately owned or whether they're vended, uh, such as an FIS or whether you actually start to leverage the public clouds on a global basis. So, you know, it's a challenge I think, to get to that point. At the same time, it's being able to do that, but concurrently managing your risk, your regulations and your ongoing services. So ironing out those manual inefficiencies concurrently to satisfy those challenges, I think is where I see the, the market right now. And technology is absolutely coming through more and more as the key tool of this.

ERIN DANGLER: Yeah, it sounds like an entire paradigm shift. It reminds me of back in the day when I went to the doctor and it used to be, I had a patient folder that they wrote down the information and then when they switched to digital format and what, and how, what a pain point that was for medical Institute. I can only imagine how it is to kind of keep the financial institutions, keep the machine up and running while you're changing the machine at the same time, which kind of leads us into our next topic is talking about disruptors in addition to all of this change in technology. So what do you think some of the biggest disruptors to the capital markets are right now?

TONY WARREN: I think this is where it gets exciting. I think at the moment there's, disruption through basic RPA, which is just a more advanced workflow and computerization of the industry, which of course, it's kind of a disruption in its own, right? In that you see more outsourcing and you're seeing more location strategies being put to play because it can actually work through these evolved infrastructures. But the, probably the biggest one that I see on the horizon is the whole tokenization. So if we think of tokenization in the same lens, as decentralized finance. Um, so essentially, it's delivering blockchain based transactions and we're seeing this today, already now with cryptocurrencies and digital assets, but ultimately, you know, is there a future that everything could effectively become decentralized and working on blockchain based transaction models. We'll probably see it as specific asset classes at the start, but you know, who knows where this could go a decade out, et cetera. So to me, this is certainly one thing that the market has got to have a very keen eye on, and we're already seeing cryptocurrency and digital assets as an ask in the mainstream. So, uh, for traditional platforms of the industry being at least compatible that you can value them, you can trade them and ultimately, you'd be able to use them as a currency.

ERIN DANGLER: So let's just take one quick step back. Can you just sort of give us a reader's digest version for some of our listeners out there who may not know what tokenization is?

TONY WARREN: Yeah, I mean, it's essentially, it is, it is full digitalization of an asset. So instead of having like the traditional methodology of currencies, uh, securities, cash, et cetera. The whole thing becomes a, um, a digital element, which can then be moved through these systems totally controlled through, um, what is termed as a smart contract. So you have like keys that you can unlock that piece of data, but essentially it is, it is a very complex, uh, code, but we're certainly on, as I said, we're on that conveyor belt, right now.

ERIN DANGLER: And you also mentioned that it's now becoming an ask for customers. They're wanting this, but maybe financial institutions aren't quite so ready for it. There seems to be a little bit of a dichotomy there. What are you seeing as far as attitudes among institutions towards cryptocurrency and digital ads?

TONY WARREN: It's almost come like a tidal wave. You knew it was out there. And now it's starting to hit the shore. You're right. I think the traditional industry is obviously still king, but more and more now there is a request to be able to, um, handle crypto currency or derivatives. So I it's clearly going to start in the, in the alternative space, but as I said before, you know, who knows where it'll go? I think we'll start to see it initially as kind of an investment rail. Then we'll have a on the back of that, the crypto OTC, but then ultimately, I think, you know, we'll see it in commercial lending. We're going to see it in corporate treasury and payment management. We're going to see it in insurance and we're going to see it in reg take.

ERIN DANGLER: What are some initial steps that need to take place in order to bring us towards this tokenized future, whether it's on the consumer side or financial institution side.

ED ABBO: If I could jump in here, if you take a big step back, there's a series of technologies and, uh, certainly, um, you know, Tony's been talking about blockchain, but if you think about it more broadly the concept of cheaper computing often referred to in lay terms as cloud computing, the concept of a low latency networks to enable faster trading sensors, to you have a dramatically exponentially, more information to process and a blockchain, et cetera. Industry and financial institutions in particular have never seen such numerous number of technology vectors change on them at one point in time. So what that means is massive innovation is possible. And with that massive disruption. So I would say that financial institutions need to adopt these technologies, embrace them and basically deliver, uh, because the alternative will mature. It'll, um, it'll be an option for customers. And so this is the digital transformation electronification that, that they need to embrace.

ERIN DANGLER: And you think an industry-wide uptake of this digitization, how could it affect the capital markets on both the buy and sell side?

TONY WARREN: It would turn the industry on its head if it really took hold because there, it would become a replacement technology end to end. And so everything would be done potentially very differently, but of course, things like regulation would have to catch up and you, that there's so much regulation in this industry and everything from uh, AML and you know, compliance monitoring would still need to wrap around it. So this is, you know, this is, that kind of envision is, is a few years off, but I don't see why things like stable coin couldn't become the first step along this journey. Probably within this decade.

ERIN DANGLER: Cryptocurrencies and tokenization are certainly a hot topic right now. But with issues like regulation and infrastructure adaptation, the capital markets probably won't be ready to start the widespread tokenization of assets for some years. However, there are other technological disruptors already in play that could be making a difference in the much more immediate future. Well, I think you may have answered my next question. You know, when you're talking, Ed I believe you said, uh, that this is a massive innovation, a massive disruption, you know, you're thinking a long, long time. Do you think we could see this within the next decade? Do you have any predictions for how long it could take?

ED ABBO: Yeah, no, I think it's happening now. And so one of the technologies is algorithms. Algorithms have been around forever. What's different now is the, uh, cheap processors, whether it's a CPU's or GPU's, um, to basically rapidly execute algorithms in real time or near real time. And that allows financial institutions and capital markets to better serve their customers. It used to be that the trades would be looked at receipts and payments would be looked at an aggregate, but with cheaper computing and algorithms, you can now analyze every single micro movement. if you will, of cash. And so with that, you can see who's being paid, who's paying, and then track that down the, uh, the supply chain, if you will, and basically make better decisions on behalf of your customer to lend them more money, because he can see that there's liquidity that's flowing through to them. And I think that that is today. You can do that with technology by integrating data across. the banking systems, uh, within a bank by integrating data that's available in the open market, sometimes referred to as open-source data and then unleashing algorithms against those unified data to make a better predictions of whether you should be lending. Or not as the case might be. And, you know, looking at a global trade, all of that can inform better servicing the customer. And you can do that today. You don't have to wait a decade to do that in the area of, of blockchain. I think the technology does need to evolve to keep up with the volume. So if you look at, you know, a reasonable stock exchange, it might be trading three million, it might be handling 3 million messages a second. I think there's a lot more overhead on blockchain. And so that's going to take a while for it to deal with those volumes, but over time that'll happen. But the bottom line, Erin is there are innovations that banks should be adopting today so they can remain competitive or actually drive their competitive advantage in the market. So, uh, so there's, this is now, it's not next year. It's not a decade from now.

ERIN DANGLER: Well, in one of those changes that we see happening now is the use of AI that's been happening in the way more recent past. Where are you seeing AI currently being used on the buy and sell side?

ED ABBO: That's really remarkable. This is really about informing people who make decisions to make better decisions. So without AI, the decision framework that most people have is relatively simple. I.E, they're looking at, um, a few attributes if you will, to make a lending decision or to make a trading decision. And often there's delay associated with exceptions. And so should I make this loan while I got to get through these checks and processes before I can make a decision. And so what we're talking about with AI and ML is being able to aggregate data from the various systems within a financial institution. So you can actually see the position of the client with the bank in real time or near real time. And then that basically informs decisions about trading and lending, and that's where the AI and the machine learning comes in, which is to basically detect would these micro-payments for example, there's a trend that this customer is likely to leave the bank. And so this is a customer attention use of AI and machine learning by studying every single change. So a corporate client that might have hundreds of millions with a bank. If you look at each payment, it might be 10 million, 5 million, et cetera. And so none of that's alarming, but over time you can have an early warning indicator through machine learning algorithms. When you see the rate of change or the frequency of payment slow down or accelerate, that might be an indicator of what the customer is intending to do. So, um, it's being applied, I should say, to improve customer engagement. Um, it's being applied in trading. It’s basically being applied, uh, to better inform decisions, better inform pricing and to better service customers. And that becomes a competitive differentiator for the institutions that are ahead deploying these technologies.

ERIN DANGLER: So this, the AI and ML can help interpret data for the financial institutions and help keep customers engaged in this 24/7 world. And Tony had mentioned, I think you both had mentioned just how regulated this industry is. How can AI help financial institutions keep up with the regulations?

TONY WARREN: I think one thing here is as the regulations grow. And as I said at the start, you know, we're an environment where the traditional departments are now disappearing, and it's really sort of an end-to-end evolution of infrastructures. And ultimately what that means is we get open APIs, we get more exposure to data, is you are increasing your data load, like a hundred-fold to you would have had in, in traditionally. And so we absolutely need tools, leveraging artificial intelligence, which would allow us to synthesize the data and also allow different sources of data to be joined to the traditional data. One of the regulations of the future that we're foreseeing is for ESG and ESG compliance and checking on portfolios. And so being able to use tools, automated tools that can synthesize, join and then create new data sets, which you can then expose for different areas of regulation we see as being increasingly important, uh, to enable you to get through this and other good example, which we've launched a partnership with our colleagues at C3 AI is the whole anti money laundering uh, process. You know, where, where this is really strong is, is as well as pulling in the SARS, it's then getting rid of false positives. And we believe that we can add correct me if I'm wrong, but I think we can reduce those false positives by applying artificial intelligence by some 80%. So it becomes efficient and doable once again, without really disrupting your operation.

ED ABBO: That's right. Um, so using, um, AI and machine learning Erin to identify financial crimes or anti money laundering, we've demonstrated that not only can you reduce what are false positive alerts as you monitor transactions. The last generation of systems had rules. And so if you triggered those roles, that would be, what's referred to as an alert, the trouble is that the rules are very course-grained and they operate on very, uh, sparse data and attributes. And so using AI and machine learning, you can actually take into account kind of an order of magnitude more factors and identifying crime. And so you can reduce the number of false alerts that the analysts have to triage. And so today, That means that, um, banks can reduce the, um, the number of people that are basically what I call fetching rocks. They're looking at these alerts and they disposition them as false. Um, so 99% of the alerts are actually false, so we can reduce the number of false alerts by 80, 85%. And we can actually identify money laundering three times more effectively than these rules. So you're not only more efficient, but you're much more effective using machine learning. And so that applies in anti-money laundering. It applies in trade surveillance. It applies in a number of different areas where you're basically applying algorithms and training algorithms on prior financial crimes to identify new crimes to really can be applied both on the customer side, the trading side, as well as on the, uh, to comply with the regulation much more effectively.

ERIN DANGLER: And I'm assuming that's, um, kind of all ties in, in general with cybersecurity.

ED ABBO: Cyber is, uh, as a massive area.

TONY WARREN: The whole thing is a whole other ball game.

ERIN DANGLER: I was going to say because it's a whole other conversation.

ED ABBO: But in a soundbite, um, Erin, the cyber area has typically, or traditionally been one where again, we're looking at signatures of traffic and comparing that signature to prior known vulnerabilities or, uh, uh, approaches to, to bypassing, um, you know, secure firewalls, et cetera. And that's another area where AI and ML can do a much more effective job because it's not looking at just prior approaches it's looking for what's referred to in the industry is anomalous traffic and network traffic if you will. And identify what the industry calls a zero-day vulnerabilities, things that you've never seen before. And it's a perfect place to apply AI and machine learning as to improve security and cybersecurity.

ERIN DANGLER: AI, advanced algorithms and machine learning have the power to greatly improve efficiency and accuracy across a range of functions within financial Institutions. Being just as adept at detecting fraud, as they are, keeping up to date with compliance issues, these technologies free up workforces to focus on the more important issues like escalating cases to fraud departments or engaging with clients. But these technologies don't just have the power to improve operational efficiencies. They also stand to improve the client experience by reconciling various data points, to help institutions understand and better serve their customers. Let's move into the future. So we've talked a lot about AI, ML, tokenization. So looking down the line, even though we know it's already here, what are ways that AI, you think in the future can help institutions better understand their customers or in what ways can it help those working in the buy and sell side? What do you see? Months, years down the line.

ED ABBO: Financial institutions have a plethora of information about their clients, but they haven't been able to use it to effectively, better serve clients. And so if you look at the emerging technologies that are now available or have the, you know, kinda been available over the last decade. So again, cheaper computing algorithms that use those, the ability to aggregate data from within a financial institution, as well as external data from outside. We can now leverage all that information to better serve the clients and the customers. And so there's a, this really a step function change in the ability to service the customer. And what does that really mean? Basically proves their liquidity, proves their working capital. That's really what clients want. And so we can now, uh, take advantage of, uh, cheaper computing cloud computing and take advantage of these AI and ML and optimization algorithms. We can take advantage of quantum computing. We can take advantage of blockchain to more effectively service the client's need. And that's really what it's all about in the capital markets and basically analyze all these data to better serve and serve clients faster. So this, the idea that it takes three days to settle something is really, um, needs to change. We need to be able to understand what's the customer intent from their payments flows and cash flows. We need to understand, you know, KYC or know your customer. We need to understand that from the day they sign up with us and continuously through the days that they're actually transacting with you. So it's not a, it's not something that happens when you sign up a client, it happens continuously and you're better serving them. So that's directionally where financial institutions need to go. Those that get there fast, have a competitive advantage in the market. Those that are laggards will not exist. And this is what digital transformation is really all about.

TONY WARREN: I agree with that. And I think very specifically, if we look at the capital markets industry, it's very conservative by nature and it's very risk adverse. And so everything is, you know, there's a lot of error handling, a lot of checking as the data moves through its process flow to create product on a daily intraday, et cetera basis. And so I think almost like the first immediate stage here is that there are tools today, which are widely used as part of the RPA evolution. So automated exception management sort of bottom-up checking of data, but then you've still got that human intervention to come in and make the repair, which is still driving that human touch across the industry. And I think therein lies a massive opportunity for technology and the, you can use machine learning. So as, as exceptions are being handled, repaired, areas are being fixed. It's done via the artificial intelligence machinery. So it's learning as you go along. So after they see this same condition for the fifth or sixth time, it now knows exactly what to do, and they can go in and make all of the repairs thus reducing the more inefficient processes. And I think we'll see that step-by-step across all areas of the buy and sell side, and that is absolutely happening now.

ERIN DANGLER: So as we wrap up today's episode, I'd love to give you both a chance just to kind of chime in your, your wisdom and advice to those listening. So someone who's listening works in a financial institution. What advice would you give them? What would you say to them about innovation right now?

ED ABBO: This isn't just technology. As Tony mentioned, this is really the difficult part, we solved the technology um, part of this. Cloud computing is now pretty prevalent, AI and machine learning and algorithms, and the ability to kind of establish teams of quants or data scientists is well known. The hard part here Erin is really, uh, changing the way that organizations do business and change comes from the top. There needs to be, um, chief executive, uh, officer's COO engagement in this. This is not something that gets delegated to the chief information officer at a bank. They play a role in it, but, uh, it really comes from the CEO or the COO of the organization. And, um, they just basically need to, uh, drive a change through the organization, because processes will change what people do, how they're compensated, all of that will effectively get changed if they're going to make it through or survive this, uh, digital disruption. So, uh, the technology is viable. It works. It's proven. And, uh, yeah, there is some refinement that needs to happen on blockchain, but cloud computing, machine learning, AI has been applied in banking and it has been done so successfully. And so now this is a change management and leadership challenge.

ERIN DANGLER: Interesting perspective, a change in leadership challenge. Because nobody likes change. We have to reign in our anxiety and these risk averse situations. So it takes strong leadership to initiate it. Very interesting. Tony, what about you?

TONY WARREN: Very similar theme. And we're, we're already seeing it in the industry that it is, it is step at a time though, to me, it is, it is an evolution because as I said, this industry will always remain as quite risk averse. So it is a step at a time, and you build up the confidence and you go to the next step, but we're already seeing now more of a culture that firms are now evolving their infrastructures. They are taking down the silos of the different departments. I mean, you know, I think we'll see, going back to our point of cryptocurrency, stable coins, ultimately tokenization and decentralized finance. That will be, a evolutionary journey kind of almost like asset class by asset class, as it cuts through the industry, but an extremely exciting time, certainly to be a tech provider for financial services, certainly more and more than I've ever known. It really is. It really is incredible, the opportunity and the pace that things are moving.

ED ABBO: I would echo what Tony said, which is you got, you just got to get started. You got to start it in one area. Maybe it's in financial crimes. Maybe it's on the trading side to apply these cloud and AI, ML capabilities. And then from there with the initial success, you expand them extrapolate and accelerate. So, uh, get started.

ERIN DANGLER: Tony Warren is executive vice president and head of strategy and solutions management at FIS. And Ed Abbo is president and CTO at C3 AI. That's it for today's show. Thanks for joining us. We'll see you next time for the season finale as we explore the rapidly changing world of compliance and risk.