Speaker 1:
From the library of the New York Stock Exchange at the corner of Wall and Broad Streets in New York City, you're Inside the ICE House, our podcast from Intercontinental Exchange on markets, leadership and vision and global business, the dream drivers that have made the NYSC an indispensable institution of global growth for over 225 years. Each week, we feature stories of those who hatch plans, create jobs and harness the engine of capitalism right here, right now at the NYSC and at ICE's exchanges and clearinghouses around the world. Now welcome Inside the ICE House, here's your host, Josh King of Intercontinental Exchange.
Josh King:
In the first half of the 20th century, the notion of artificially- intelligent robots was introduced to the world through the movies. Maybe it was a young Judy Garland who deserves some credit marching down the Yellow Brick Road in 1939 alongside Dorothy's dog, Toto, and a host of characters, including a heartless yet fully functional robot, the Tin Man in The Wizard of Oz. By the 1950s, the notion of artificial intelligence had become deeply ingrained in the minds of the era's foremost intellectuals. Among them was Alan Turing, often hailed as the father of theoretical computer science created by Benedict Cumberbatch in 2014's The Imitation game. Turing was one of the pioneers who dared pose the question, could machines think?
Alan Turing:
Could machines ever think as human beings do? Most people say no. You are not most people. Well, the problem is you are asking a stupid question.
Speaker 4:
I am?
Alan Turing:
Of course, machines can't think as people do. The machine is different from a person, hence they think differently. The interesting question is just because something thinks differently from you, does that mean it's not thinking?
Josh King:
Yeah, I can't do any proper Benedict, so that wasn't me. That was the real guy. But in the past decade, the capacity of computers to think as conceived by Turing has exploded. Whether it was Apple's unveiling of Siri on the iPhone 4S in 2011, Amazon's introduction of Alexa in 2014 or OpenAI's release of ChatGPT less than two years ago, the aptitude of a machine to respond to everyday inquiries or even engaging conversation has been steadily increasing. Welcoming the evolution of AI and its influence on both the workplace and product development is today's guest, Steve Hasker, president and CEO of Thomson Reuters. That's NYSE took a symbol TRI. Since taking the helm in 2020, Steve has helped the company become a forerunner in AI.
That means not only driving innovation within its own tech stack, but also strategically leveraging investments and acquisitions to integrate AI into the suite of products used by Thomson Reuters clients. In a minute, Steve and I are going to dive into Thomson Reuters' implementation of artificial intelligence and how it's impacting various segments of the business. We're going to discuss the excitement and trepidation among employees that comes with AI and hit on how it's shaping Thomson Reuters near and long term. Our conversation with Steve Hasker, president and CEO of Thomson Reuters, it's coming up right after this.
Speaker 5:
Explore the world the Viking way from the quiet comfort of elegant small ships with no children and no casinos. We actually have reinvented ocean voyages, designing all-inclusive experiences for the thinking person. Viking, voted world's best by both travel and Leisure and Condé Nast Traveler. Learn more at viking.com.
Josh King:
Welcome back. Please remember to subscribe wherever you listen and rate and review us on Apple Podcasts so other folks know where to find the show. Our guest today, Steve Hasker, is the president and CEO of Thomson Reuters. That's NYC ticker symbol TRI. With roots in Australia and a career spanning North America, Europe and Asia, Steve brings over 30 years of experience with previous stops, including McKinsey, Nielsen, the Creative Artists Agency among other places. Since taking the helm in March 2020, Steve has led Thomson Reuters and its team of close to 26,000 employees. Steve, thanks so much for joining us Inside the ICE House. It's a pleasure to have you here at the New York Stock Exchange.
Steve Hasker:
Thanks for having me, Josh. Good to be here.
Josh King:
When people hear Thomson Reuters, Steve, they often think of Reuters News and its renowned global coverage since Paul Reuter established a newswire agency at the London Royal Exchange in 1851. I worked with many Reuters correspondents and photographers when I worked at the White House in the 1990s. Here at ICE and the NYSC, we're constantly working with Reuters journalists who cover the broader financial markets, but as you said in a recent interview with Andy Serwer of Barron's, we are a tech stock, not a media stock. Given that Reuters News is just a small part of the company, what makes up most of the bulk of the firm?
Steve Hasker:
Yeah, so Reuters News is just under 10% of our revenues. The rest of the revenues are within what we call the big three, which serve professionals, so legal professionals, tax and accounting professionals, risk fraud compliance, government agencies, court systems and so forth. What we provide to those professionals is content-driven technology, so a combination of unique and content that we've built up over decades with increasing amounts of AI. We've been in the advanced machine learning business for 30 plus years, but obviously, generative AI is the next step delivered through what we aspire to make best-of-breed software.
Josh King:
You've been leading Thomson Reuters as its president and CEO since March 2020 as I talked about in the introduction. But, Steve, your career really began back in the 1990s at accounting firm PWC, I would add to the bio, followed by being a partner at McKinsey. Paint us a picture of how one goes from accounting to consulting to eventually leading this New York Stock Exchange listed enterprise.
Steve Hasker:
Yeah, I started as many people do in public accounting, which, for me, was a great grounding and training at what was then PW, now PWC's Melbourne office. The great privilege that I was able to benefit from was that the partner group at PW in Melbourne was a fantastic group of professionals who upheld the highest standards and at the same time were a very collegial and caring bunch of people. So for me, it was a great way to start my career. I spent four and a bit years there and then went to Columbia University, did an MBA and an international affairs masters, which was one of the best experiences of my life. Out of that, I was recruited into McKinsey. I had a chance to work at McKinsey both here in New York and also in Australia amongst other places, and ended up being there for 11 years, which is longer than most people stay at McKinsey, but I thoroughly enjoyed it.
Josh King:
From Melbourne to New York, now you're in Toronto and your time really spanning the globe, Steve. How has global experience that you've had helped you better understand the global customer and consumer that many parts of Thomson Reuters supports?
Steve Hasker:
Growing up, one of the things my mother always said was, "You never quite understand until you experience it yourself." So I've been lucky enough to spend time first starting my career in Australia, but significant time in China and India, in Portugal, in the UK, of course, the U.S. and Canada. I think that's given me a global perspective. I don't know whether it's a uniquely global perspective, but I certainly feel like when I travel, as I often do and visit our offices and our customers in different parts of the world, I can relate to the challenges and opportunities that they describe in a way that I really don't think I could had I not spent significant time outside of the United States.
Josh King:
What's your method of keeping track of everything? If you're sitting up in Toronto and understand how the global machinery of Thomson Reuters is operating, when you aren't able to get on a plane and go visit with clients, how do you know that you're doing a decent job keeping your eye on everything?
Steve Hasker:
Yeah. Well, I think our board keep me pretty honest in terms of whether I'm doing a decent job or not. As to my colleagues, we've tried to create an environment that's very flat and non-hierarchical where people feel empowered to speak up. We're not perfect on that dimension, but I certainly hear it from colleagues when they think there's something I could do to improve. There's a few things that I do to try to keep up to speed. The first is I am an avid reader of news. I have all sorts of news sources that I go to as I know you do every single day, and I rarely miss the opportunity to spend as much time as possible reading the news, obviously starting with Reuters and reuters.com. So that's the first one.
The second one is just what I would like to think is an insatiable appetite to interact with customers. So no matter where in the world, I'm always looking for an opportunity to listen to our customers and understand what opportunities and challenges they face. I try to do more listening than talking in those meetings, more listening than selling, but I always find I learn something from one of our customers, whether that's an existing customer or a prospect. Then last but not least, I have a bit of a rule, well, within that I have a rule that I'd like to speak with at least one customer every single day. I try to pursue that as I work my way through a work week. Also, I have a-
Josh King:
Do you manage that rule yourself or do you have someone who serves-
Steve Hasker:
No, I manage that myself. I just do my best to make sure that as I look at the calendar coming into a week, there is an abundance of customer interactions. Then thirdly, what I say to our colleagues is, "If you've got a question or an issue or feedback, suggestion, just give me a call or send me an email or a Teams message or some other form of communication. Don't stew on it, and don't feel like you can't or shouldn't raise that issue." I try to respond within that day to any incoming suggestion or comment or question from a colleague. I don't know that I'm perfect at that, but I think it certainly, it keeps me aware of what our frontline folks are facing. I think that's really important.
Josh King:
Before we dive into the specific segments of Thomson Reuters' business and talk about the integration of AI into it and your products, I want to just rewind the clock a little bit. You conducted the first quarter earnings for Thomson Reuters in May. The call highlighted in those results were an increase in 9% in organic revenue, 8% total company revenue compared to the fourth quarter last year. How have these results influenced the company's outlook for the rest of '24? Any expectations to raise expectations going forward?
Steve Hasker:
Well, we raised our revenue guidance slightly coming out of the first quarter because we did have a good first quarter. But I'm always very cautious and not to over-celebrate a good quarter and think that it's necessarily going to parlay into the rest of the year. We've got a lot of work to do in serving our customers and improving that proposition both in terms of the products, the support and the after-sales service. So we are laser focused on that, and if we do that, then the results will follow. But certainly, as I've said to our team, better to make a good start than the opposite, but it doesn't count for much. What happens in the next few quarters will matter.
Josh King:
You mentioned the big three segments earlier in our conversation, legal professionals, corporates and tax and accounting pros. Collectively, that accounts for about 83% of total revenues for the first quarter. Organic revenues, these segments also grew 10%, which you said on the call was an all-time high. What factors do you think drove the increase in, and do you anticipate those highs continuing through the rest of the year?
Steve Hasker:
Well, I think there's a couple of things that we consider to be tailwinds. The first one is the rising complexity associated with compliance. So whether that's legal compliance, tax compliance, risk ESG, internal governance and so forth, what we see across the world and across every segment of corporate customer is increasing rules and regulations. Corporations and the advisors that serve them and the government agencies that obviously support those regulations cannot cope with that increasing complexity by merely adding more people. It's just not scalable or sustainable. So they have to turn to content-driven technology to help them navigate that, and we think we're one of the few providers that can really lean in and support them in that process. So that's the first thing. The second thing is, as I mentioned before, we've been in the AI and advanced machine learning business for 30 years.
We put a functioning search algorithm on the front of Westlaw, which is our legal research database back in the early 1990s. So really before any of us were thinking about Google and search and so forth, we were in the search business and certainly created a functioning search algorithm. So we've got a pedigree in history there. As generative AI has come along, so with obviously the release of ChatGPT GPT-4 18 months ago, that has created, I think, a real sense of excitement amongst our customers as to what this could mean starting in legal, but increasingly in tax and accounting and in risk fraud and compliance and other areas. It's also injected a lot of energy internally. I have engineers coming up to me and saying, "Steve, I just worked a weekend not because I needed to, I had to because I wanted to." The stuff we are inventing, the problems we are trying to solve are really cool for our top talent, and that makes things pretty exciting.
Josh King:
That does make things exciting. Another exciting part of your business is when you send correspondence to far-flung places around the world to report the latest news. Reuters News' organic growth grew 17%, which stands out maybe in a media landscape where a lot of companies are facing mass layoffs and missed revenue targets. How has Reuters News achieved the growth that it has during these turbulent times?
Steve Hasker:
Well, the first thing to note in that particular result was some non-recurring revenues that stem from providing access to the Reuters News file to a couple of the large language model providers, so we don't think that will recur. So that was a little bit of an anomaly. But having said that, the biggest customer to Reuters is the London Stock Exchange Group. So we have a news agreement with LSEG where we provide the news into their financial data business that used to be owned by TR, and that's a 30-year agreement. We're a couple of years into it, and we're really focused on strengthening the partnership with LSEG and their team and also helping them better compete globally and meet their growth aspirations. So that's a big part of it. Reuters is 2,600 journalists spread across the world.
In many countries, it's the last purely independent fact-based news service standing. So I think it's a very important role for humanity. Equally importantly, it's the largest wholesaler of news in the world. So when you switch on the broadcast news at night or go to a website or read a newspaper, often the text and the story, the facts and the data, the photographs or the video will be from Reuters, and we're very proud of that role. The news landscape has been challenged by the advent of digital. It will be challenged again with the onset of generative AI, but we think the fact that Reuters sits alongside the big three as we call them and alongside the very significant investments we are making in generative AI, we think that gives Reuters an advantage in navigating the next wave of disruption.
Josh King:
Reuters represents, as you said at the beginning of our conversation, only about 10% of your total revenue, which is, in some ways, like the challenge we face at ICE Intercontinental Exchange. New York Stock Exchange, our best-known company is a similar percentage of the total revenue of the business, and yet that's what people always know us as. They say, "NYSE Parent ICE does this or does that," so I share some of your challenge there. But despite its small size compared to the other segments, Reuters News remains arguably the most well-known part of your business. How do you ensure that the brand Reuters News that what it's created is upheld while adapting to the new ways that people are consuming media?
Steve Hasker:
Yeah, this is where I think Reuters is truly differentiated. In 1941, the Reuters board established what was then called the Trust Agreement, now, the Trust Principles that basically in essence require and mandate that any information published by Reuters is independent fact-based, triple-checked and not subject to bias. So when you go to reuters.com and you read a story, what you won't see is a lean to the left or a lean to the right or a narrative that involves a lot of opinion. You'll see independent fact-based news that has been triple- checked and verified, and those Trust Principles were created in 1941 governed Reuters through to Thomson's acquisition of Reuters in or around 2008. As part of that deal or that acquisition, the Trust Principles apply to all of Thomson Reuters, and we consider it a bit of a superpower. Obviously, it requires significant investment in systems and in people and in governance to make sure that we uphold those values, but particularly in a world of increasing misinformation and skepticism about news, we are finding it to be an increasingly important and differentiating set of principles for us.
Josh King:
I don't think Reuters was on the immediate front line of this, but I'm pretty sure you've got into this fairly quickly and reacted as other news organizations did. But when you saw news break about Catherine, Princess of Wales and the retraction of the photo of her family by whatever wire put her photo out first, how did that land on you?
Steve Hasker:
Well, much more experienced and clever people than I make those calls, so specifically, Alessandra Galloni, editor-in-chief and Alex Friedman, who overlooks the application of the principles on everything we publish. So we have a set of process and protocols that when an event like that happens, whether it's a photograph that we have taken or that we have distributed, we'll have a set of processes that we go through to verify that it is an authentic photo and that we can stand behind it. That pertains to all content that we publish. Obviously, that's a significant challenge, but again, it is an opportunity for us to differentiate ourselves. I think if you look back, certainly if I look back at my time in the company, but more importantly, in 175 years of Reuters history, we've done that as well, if not better than anyone else.
Josh King:
When Lance and I work on having a conversation with someone like you, we used to run a simple search and begin to do our reading. Now we've got generative AI beginning our work for us at the top line. That is a transformative challenge for journalists who do the same thing. A concern across journalism today about the impact that artificial intelligence could have on it, so far, how has AI impacted your newsrooms?
Steve Hasker:
Yeah, the way I've been talking about AI now for a couple of years and I think is true and will prove out over the next few years, is that in the course of my career, I saw the advent of a PC on every desk. When I first started in public accounting, it was a pen and paper, and then my first, second year in all of a sudden, we're all given PCs. We then had the internet. We had mobile, social, cloud. In my view, generative AI will be bigger than any one of those in terms of its transformation of the lives of professionals. The ability to have a lot of the early foundational, often mundane, repetitive work performed by machines and automated and allow the human beings and the professionals and the practitioners to apply their judgment and their networks and relationships, their narrative and storytelling on the top of a set of facts I think is transformative for the legal profession, for the tax and accounting profession, for the risk fraud and compliance and equally importantly, news, to your question.
So what Alessandra has done is to have a hard look at everything that Thomson Reuters is doing in the generative AI space, the internal platforms we've built, the access we are providing to large language models, the governance and controls that we're putting across all of this, and she's basically going through and picking and choosing that which she's going to apply to our newsroom. That process is well underway. She's taken one of her biggest and best talents, Jane Barrett and injected Jane into our Thomson Reuters labs to learn everything that's going on in terms of the early experimentation we're doing, and Jane will bring that expertise back. So I'm confident that what we're going to see is the role of a journalist become more productive, importantly, more fun, and in a sense, an opportunity for journalists to add more value.
So in other words, instead of doing that the early, for one of a better term, grunt work in preparing a storyline, it really is the machines can do a first version. Then the question is how do the journalists and the editors step in and really take it to the next level? To us, that feels like a big opportunity. Of course, one of the reactions is, "Well, this is going to steal jobs." The way we think about it is to say, "Look, your job won't be taken away by AI, but it might be taken away by someone who's using AI." I know we've all heard that adage, but that really rings true, I think, across Thomson Reuters, not only within Reuters, firstly. Secondly, as Alessandra will say, if I just look at the last 12 months, Reuters has produced incredible coverage of two major conflicts. Most recently, we were awarded Prize for our photographs in Gaza.
In addition to that, we are obviously covering arguably one of the greatest election years in history with so many countries going through the electoral process. We're facing this very interesting interest rate climate in speculation as to what the various central banks will do. We've got demographic shifts, we've got climate change, we've got AI, Reuters is covering all of those stories. So one of the conversations that Alessandra and I have with Paul Bascobert, who's the president of Reuters is, "How can we use AI to cover more and increase the capacity of our newsrooms and our bureaus across the world?" That's a conversation around expansion and growth. That's not a conversation around contraction and headcount replacement, and my bet is that's the way it will stay, at least for us at Reuters.
Josh King:
One is tempted as we move from the newsroom to the broader workplace, and as I reflect on the 15 minutes of our conversation so far to say Reuters stands alone and the best in its application of artificial intelligence, not only in news, but in the three other major parts of your business. Yet, if we're all being honest with each other, you've got competitors who are working hard to do a lot of the same thing. With so many companies investing resources into AI and developing proprietary models on their own, how does Thomson Reuters' approach to AI distinguish itself from its competitors that are also embracing the innovations?
Steve Hasker:
I think there's a couple of areas where at least in the first innings of the generative AI revolution, there's a couple of places where we see competitive advantages. The first is in our unique and proprietary content. The fact of the matter is that the large language models need to be trained, and they need to have unique and proprietary data sets that are correct and that have deep domain expertise applied to them in order to be valuable to a professional. So we have as much if not more of that unique data in legal, in tax, in risk and in news amongst other areas as anybody on the planet. So we think that's a competitive advantage as we navigate through. The second is one of the terms you used is prompt engineering.
But essentially, it's one thing to put a question to a large language model. It's another thing to understand how a professional thinks. When a professional, be it a lawyer or a tax accounting professional or whomever asks a particular question, they type it in or they use voice activation, to interpret what they really want and how that should be translated to interact with a model is no mean feat. The better you do that, the more accurate the results and therefore, the more value you're providing to that professional. We have many thousands of qualified lawyers and public accountants in our 26,000 colleagues. So what we're trying to do is really take all of that intellectual capital and all of that deep expertise about those professions and make sure it's applied in each and every one of our products. I think so far, the evidence suggests that David Wong and our product team are doing a very good job of that, but obviously, it's a constant work in progress.
Josh King:
That's how you are using your in-house team to grow organically, and yet, there's other stuff out there that you can buy and bolt on and continue to build what we all know as Thomson Reuters. An interview that you did with the FT in March highlighted an influx of, I think, $8 billion that Thomson Reuters had to spend on AI-related investments. If we're going to cover some of those beginning in 2023, acquisitions have included SurePrep, Casetext, Pagero among others. When these deals were finalized and a lot of firepower was brought to bear after making these additions, what's the process in evaluating and determining the right opportunities to move forward with and new deals to make?
Steve Hasker:
Yeah, so as you said, Josh, we've spent about $2 billion over the last 18 months on a handful of acquisitions with real focus on our customers. So we look at particularly the big three, and to some extent, Reuters as well, and try to understand what is it that we need to do in terms of organic build, and then how can we supplement that inorganically to enhance the experience of those customers to provide them with better and faster access to information, better decision-making tools, a better end-to-end customer experience? Each of the acquisitions you mentioned are very much on point for that. So what we haven't done is look to create an entirely new business or entirely new vertical. We've very much stuck to our knitting. That's the first lens that we apply.
Josh King:
So when I throw around names SurePrep, Casetext and Pagero, what do those things do?
Steve Hasker:
Yeah, so SurePrep is, we think, the leading AI-driven document ingestion engine for the tax preparation process. So think about your tax accounting professional and all the myriad of documents that he or she has to come to terms with in order to populate a first version of the return, SurePrep automates that using AI. So that's the first one. The second one that we did was Casetext. What Casetext had done is got early access to ChatGPT-4 and we think cracked the code in terms of how to put together accessing a large language model with unique and proprietary data. What Jake Heller and his team had done, I think, was really significant in translating that into an AI legal assistant. So we're excited about taking CoCounsel and applying it across our entire product set and our entire business. So that one was very focused on generative AI.
Pagero, which when we bought it was a publicly-listed company in Sweden, and we only closed that deal a few months ago, but Pagero is the world's leading single platform for what's called e-invoicing. So this is the calculation of indirect tax or sales and excise tax and e-invoicing is a process by which a number of governments, particularly in Europe, increasingly Southeast, Asia and Latin America have mandated the automatic digital remittance of that tax upon completion of the transaction. The Pagero team under Bengt Nilsson had cracked the code in terms of the software and underpinnings of that, so we're excited about bringing that into the fold as well. It's early days. So those are the kinds of things we go after. We look firstly at the customer experience, is it additive to the customer experience?
We then look at the financials. We want to make sure that there's value for our shareholders, not just the exiting shareholders in any deal that we do. Then thirdly, we look at the tech debt. We don't want to acquire tech debt. Obviously, nothing's ever completely perfect, but we've done a lot of cleanup work over the last few years on our own legacy tech debt. We're in good shape, and we don't want to go backwards. Then lastly, we're looking for cultures not necessarily that are perfectly congruent with ours, but certainly where we think we're confident that the teams can come together and really create something bigger and better than either team would've done on their own. Certainly, the three deals that you mentioned we think are on track to do that.
Josh King:
Before we head to the break, Steve, beyond traditional M&A, how is Thomson Reuters leveraging the extensive data sets from Reuters News to help other AI groups train their own language models?
Steve Hasker:
Yeah, so Paul Bascobert and Alphonse Hardel at Reuters did a lot of work in the backend of last year as other media organizations have done to figure out ways in which these models could be trained with the best information. So in other words, one of the things that Paul said early on was, "Look, these models need to be trained, and they may as well be trained by the best, and the best we believe is the Reuters news file." So he was able to do a number of these agreements, which a finite in period and a little bit experimental in their nature, but certainly we're proud of getting that fact-based independent journalism into those large language models to help prevent hallucinations and increase the quality of output. That's a really important step we think, not only for the models and for Reuters, but for humanity more broadly.
Josh King:
After the break, Thomson Reuters president and CEO, Steve Hasker and I are going to continue to discuss the company's focus on AI, its innovation, and what the future holds for the business. All that is more, it's coming up right after this.
Speaker 7:
To prove that there is such a thing as free money, we gave people a little sample of Ibotta.
Speaker 8:
Hi, could I interest you in a sample from Ibotta?
Speaker 9:
What's going on?
Speaker 10:
It's just money.
Speaker 8:
Ibotta is an app that gives you cash back every time you shop.
Speaker 11:
What is it?
Speaker 8:
It's money.
Speaker 10:
If I can just take this.
Speaker 12:
This is real?
Speaker 13:
This is insane.
Speaker 14:
What's the catch?
Speaker 8:
There's no catch. It's just cash.
Speaker 7:
Get cash back every time you shop with Ibotta.
Josh King:
Welcome back. If you are enjoying this conversation and want to hear more from guests like Steve Hasker, the president and CEO of Thomson Reuters, that's NYSE TRI, remember to subscribe to Inside the ICE House, our podcast, wherever you listen to your podcasts, and maybe give us a five-star rating and a review on Apple Podcasts. Before the break, Steve and I were discussing Thomson Reuters' recent first quarter earnings call, Steve's journey to leading the company and how Thomson Reuters is integrating AI. So Steve, in November, Thomson Reuters launched GenAI initiatives aimed at benefiting professionals across its big three segments. These included the introduction of AI-assisted research on your Westlaw Precision. How does this technology enhance client efficiency and facilitate their ability to navigate and resolve some of their complex legal issues?
Steve Hasker:
Yeah, so it enhances our products in a number of different ways. The first is, it supplements the proprietary information we already have, so large language models built on an enormous mass of information, much of it publicly available. When you combine that with our own unique proprietary content, you get a deeper, richer perspective on a professional's question. So that's the first. The second is the conversational aspect of accessing a large language model. So it very much turns a piece of research into the first draft of a memo that a young lawyer can put in front of a partner or a law firm can put in front of a client. Then lastly, it provides the basis through which various applications can be launched in response to a particular question. So it really is somewhat of a unifier across not only our products, but more importantly, a customer experience.
Josh King:
As AI continues its rapid evolution, becoming increasingly more impactful across all of our industries, there's this growing belief that companies have to integrate it into what they do. Arvind Krishna has been on this show before. He's the chairman and CEO of IBM, that's ticker symbol IBM if you wanted to know. He's been a big proponent of the idea of dating back to the early days of the pandemic. I want to listen to an interview that Arvind did with CNN where he talks about why the implementation of AI is so crucial.
Arvind Krishna:
I think everybody is going to adopt AI. Why? How do you do better customer service for your end customers? How do you really make your employees more productive? How do you automate operations all inside the enterprise? All of these are going to be use cases for AI to unlock that 16 trillion in global productivity we all want by the end of the decade. The same way that electricity happened at the turn of the last century and everybody went through electrification, I think AI is going to infuse every business and every enterprise in this century.
Josh King:
"AI is going to infuse every business in every enterprise in this century." You said something similar to that before the break, but do you agree with Arvind? Do you go tangential to him in any way?
Steve Hasker:
No, I do agree with Arvind. I think for us, where we've started is to inject generative AI into our customer-facing products, and that's been the exhaustive focus over the last 18 months or so. To your question, we've started to get meaningful improvements to our products and new products in the marketplace as a result of that agenda. I think the other place, though, is how do our own folks internally benefit from accessing these tools? So in other words, does it allow them to automate some of the tasks that they hate the most in their daily work? Does it get them home to see the family earlier at night? Does it free them up to think creatively and do advanced problem solving versus some of the more menial work? We see a pretty big opportunity across the entire company. So we're running experiments with our software engineers and our coders, and we're seeing meaningful upticks in their productivity and also importantly, their job satisfaction.
We're implying it to HR, our finance, our corporate legal and compliance functions, and we're starting to see the same kinds of results. What we haven't done is made any promises to Wall Street as to what that will mean in terms of productivity and margins and headcount and so forth, because I do think it's early days. I think all boards and chief executives will face that decision as to what is the magnitude and timing of those productivity benefits and what is the proportion of that that can be reinvested back in growth versus return to shareholders? So a fairly traditional corporate finance type question, but that's all in front of us. To Arvind's point about 16 trillion in additional productivity, I think it's going to be a pretty exciting few years.
Josh King:
You said earlier that your board keeps you pretty honest about the way you keep your eye on the business, and you brought the board up again now. Just like ICE as well, over the last few years, we've had considerable refreshment across a broad swath of the board. As you think with your team and your existing board about how to bring in new talent on to the board, what have some of the recent appointments done in terms of coming to you with some of these insights and specialties?
Steve Hasker:
Yeah, since I've been at the company, we've had a number of long-tenured board members retire, and it's given us the opportunity to replenish and bring in new talent. So we focused on technology talent. We focused on board members who bring in significant experience around governance, around entrepreneurship and around people management. So when I face a problem across any of those dimensions, I know and I have a great deal of confidence that I'm going to get real expertise and guidance from one or more board members through those different problems. So it's always a work in progress. It's like any management team or any board, the composition is always something you've got to review and revisit and undergo that process of constant replenishment. But certainly in my four and a bit years at TRI, it's been a good experience with the board and not only the guidance, but also the support and help that they've provided to me and the team.
Josh King:
Let's shift a little bit to the 2023 Futures Professional Study that Thomson Reuters conducted. It surveyed 1200 professionals across North America, South America and the UK. What was the motivation behind doing the survey and the specific objectives you aimed to achieve through the results that you saw?
Steve Hasker:
So the genesis of the Future of Professionals work and research was really during the pandemic where more and more of our customers, whether they were managing partners of law firms or the same in the tax and accounting practice or general councils, heads of tax CFOs within corporations, were starting to ask us and ask our account and salespeople questions like, "How should I think about hybrid and virtual and work from home?" "How should I think about attracting, retaining, developing young tech digital talent because it's new to me and I realize, 'I need to know more of that?'" "How should I think about attracting data scientists? What does that mean? Where do I go looking for them? What's my angle and pitch to get them?"
Then, of course, along came generative AI, so it started in the pandemic when I think the war for talent really heated up, and we saw initially fairly subdued turnover amongst the colleagues within our customer base, and then we saw it really spike up. I think more and more boards and chief executives, managing partners were starting to cry for help or support or some insights. We felt, hopefully without any arrogance, that given the 600,000 customers that we serve, and given it may be equally importantly, all of the information that we collect through the different parts of the business that we could harness that and provide some insights. I think we're still at the start of this journey, but it's certainly been well received by our customers in terms of some of those insights and the support we've provided.
Josh King:
Among the 1200 people that you surveyed for that study, 67% of them anticipated the emergence of AI and generative AI to have either a transformational or high-impact change on their profession within the next five years. It's 14 points higher than the second most cited impact in the profession, which is economic recession and cost of living. What insights do you derive from the response that AI has believed to have such a greater influence in the workforce than impact of things like mortgages, rents and everyday necessities for people?
Steve Hasker:
Well, I think it comes from a couple of places. I think anyone who's accessed one of the large language models, which by now is most of us, whether that's ChatGPT-4 or Claude or Llama or whichever one you tend to go to, I think you don't have to spend long with those models to know that it's going to change professional work. It's going to change the nature of research, the nature of drafting, the nature of document analysis, and therefore, the top of the house within our customers, I think, sees and recognizes that. Then it's reinforced when they go talk to the highest potential recruits. One of the first questions that recruits ask is, "Which technologies will I have access to? How are you thinking about providing access to generative AI, and what's your roadmap for that?"
I think that really emphasizes to them, "Hold on. We've got no option here." We're going to have to get immersed in this. As they play that out, they can see implications in terms of the apprenticeship model they provide, the kinds of talent they bring in, the end business model, and for the professionals, their billing and fee structures all the way back to the curricula within the graduate schools that they're drawing from. So there are very few aspects of their business that they can't see an impact. Of course, then the debate is, to what extent will that impact be unequivocally positive versus provide some real challenges? I think that debate is ongoing.
Josh King:
Positive versus providing some challenges, professionals in reacting to the survey also expressed some fears involving accuracy and data security. In the survey, Maria Apazoglou, your head of AI for Thomson Reuters expressed the importance of trust in an interview with Amazon Web Services last year. I want to listen to a little bit of what she had to say.
Maria Apazoglou:
So we wanted to make sure that as we are looking at our products to be trustworthy and as our customers are trusting us with their data that the AI that we have is trustworthy and trust embedded within the products. We start with the data because the first thing that people will want to do, or the first thing that AI needs is data. So we first look at how do we ensure that the data that's being used is ethical? Then as we move into the process of creating a model, before that model goes into production, we have mechanisms in place to look at concepts like, what are the monitoring that you will have to ensure that your model still performs the way that you want it to do?
Josh King:
Steve, with so much information at the fingertips of these AI models, how is Thomson Reuters assuring that the models can be secure and responsible?
Steve Hasker:
Yeah, Maria gave you a bit of an insight into it. The first is, we have to understand what the source information is, where it comes from, what its provenance is, how it's been framed and provided, what access rights are there? That's the first. The second is, we have to understand in depth how the algorithms are written and how the models are created and how they lead to the output. So to say, "Oh, these things are black boxes," and to leave it at that is insufficient. I think that's essentially what Maria was saying. So we've taken an attitude of getting as much visibility as we possibly can and making sure we review an audit every step of the way, because all it's going to take is for a provider somewhere to neglect one or more of those steps. Then I think the customer base-
Josh King:
Garbage in, garbage out.
Steve Hasker:
... yeah, or the broader populace will start to be skeptical. I don't think that will derail the train, but it'll certainly slow it down, and I think some of the benefits therefore will be delayed.
Josh King:
Okay. So we have now discussed some of the results of the survey. I'm curious how you think the numbers from the 2023 Future of Professional Survey to map out the near and long-term future of Thomson Reuters will play out. Do you think we're going to see, as I alluded to in the introduction, a Wizard of Oz-like robot walking freely along the hallways of your Toronto headquarters, just as we talked about earlier, the Tin Man walking down the Yellow Brick Road at Thomson Reuters? Where does this all lead, Steve?
Steve Hasker:
Robotics is hard, and I think we should distinguish between robotics and generative AI. I think robotics is on a path. It's been on a path for many decades, and it will continue to improve and move ahead. Where I think we'll see very significant progress is I think in three to five years, every single practicing attorney will have a legal AI assistant. I think that up to 1/3 or more of the legal tasks will be automated. If you apply that first to the legal profession and then to others, you're talking about a pretty significant increase in productivity. You're talking about, I hope-
Josh King:
No more time sheets.
Steve Hasker:
... a significant increase in job satisfaction and work-life balance. Because if you look at the legal profession as one example, work-life balance very significantly deteriorated. It's probably never been great. I think my practicing attorney friends would argue it's never been a fantastic work-life balance, but it very significantly deteriorated for practicing attorneys through the pandemic. It has, to some extent, improved, but it's still not great. Then when you go across the tax and accounting, there is a dearth of talent.
If you talk to any person who manages a tax and accounting practice, whether it's one of the largest global firms or whether it's a much smaller high street firm, they will tell you one of the biggest problems they face is getting young talent interested in the profession. Therefore, AI and automation has a slightly different role to play, which is to allow the practices to perform the work they need to perform because the number of tax returns is going up, the number of audits is going up, and the complexity of those returns and audits is also going up, and yet there's not the talent coming in to replace those who are retiring. So AI has a different but equally important role to play there, and I think that's what we're trying to harness and help and help promote.
Josh King:
So I think we can now pivot away from AI for a bit anyway. Steve, in 2020, shortly after your tenure as president and CEO began, Thomson Reuters announced that its plan to achieve net-zero emissions by 2050 or sooner. Four years into that plan, how has the company prioritized sustainability and work toward achieving that goal?
Steve Hasker:
Yeah, so we're ahead of track. A couple of weeks ago, we published our Social Impact Report, which frames up the work we do in the community, the work we're doing in terms of our climate footprint and our environmental goals and also all of our objectives and our progress against diversity in its various forms. On the climate front, we have the benefit, we have a print facility that manufacturers books. But other than that, we're not in the industrial space. We're not moving large numbers of people or materials around, but nevertheless, we see a very significant responsibility to move to that net neutral position as fast as possible. We are ahead of track, and we're just going to keep chipping away at that.
Josh King:
Thomson Reuters, your executive team really has a diverse composition. I think over half are either female or minority. How does having leadership with many backgrounds and experiences enhance the company's ability to act in the best interest of all employees and your clients?
Steve Hasker:
Yeah, I think it's pretty simple. The first is, we need to be a diverse team in order to reflect and represent our customers. Our customers are diverse, and if we are not, then we're going to provide an unconscious bias to our product development and the way in which we serve and support customers. So that's the first thing. I think the second thing is, in my experience, the diversity of the team adds to the richness and quality of the problem solving that that team undertakes. It's no more complicated than that. So the more diverse the team becomes, the greater variety of perspectives you get in a problem-solving process and the better outcome you get. Certainly, I think in my time at Thomson Reuters, we've seen the benefits of that, and it's a constant work in progress.
Josh King:
As we wrap up, Steve, Thomson Corporation and Reuters Group came together to form Thomson Reuters all the way back now in 2008. With the 20-year anniversary, not too far off, what do you think are the goals and aspirations for the company as it continues to evolve and reach this significant milestone?
Steve Hasker:
Yeah, we've set out an objective to be the world's leading content-driven technology company where we combine unique and proprietary data sets, AI, best-of-breed software. We'd also like to be the most innovative company in our sector, so the business information services sector. Those are lofty goals, and there's some tremendous companies and competitors in that space, but we've made real progress against those things. Going forward, we think we can play a really meaningful role in helping companies and their advisors manage this increasing complexity, and that's a tailwind for us. It's also, it's a big opportunity, but it's a challenge that our customers face, and we're looking forward to helping them meet it.
Josh King:
Looking forward to helping them meet it. Steve, thanks so much for joining us Inside the ICE House.
Steve Hasker:
Thanks, Josh.
Josh King:
That's our conversation for this week. Our guest was Steve Hasker, president and CEO of Thomson Reuters. That's NYSC ticker symbol TRI. If you like what you heard, please rate us on Apple Podcasts so other folks know where to find us. If you've got a comment or a question you'd like one of our experts to tackle on a future show or to hear from one of our CEOs like Steve Hasker, make sure to leave us a review. Email us at [email protected] or tweet at us @ICEHousePodcast. Our show was produced by Lance Glinn, with production assistance, editing and engineering from Sam Iannotti. Pete Asch is the director of programming and production at ICE. I'm Josh King, your host, signing off from the library of the New York Stock Exchange. Thanks for listening. We'll talk to you next week.
Speaker 1:
Information contained in this podcast was obtained in part from publicly-available sources and not independently verified. Neither ICE nor its affiliates make any representations or warranties expressed or implied as to the accuracy or completeness of the information and do not sponsor, approve or endorse any of the content herein, all of which is presented solely for information-only and educational purposes. Nothing herein constitutes an offer to sell a solicitation of an offer to buy any security or a recommendation of any security or trading practice. Some portions of the preceding conversation may have been edited for the purpose of length or clarity.