Transcripte: A look to the future

The fourth edition of the HEC S&O Purpose Day, hosted by the HEC Paris S&O Institute Purpose Center in partnership with the Oxford University Centre for Corporate Reputation and POLIMI Graduate School of Management. The event took place on April 7, 2026, at Maison Internationale, Cité universitaire de Paris.

One of the highlight of the day was

Panel Discussion

Purpose & Leadership at the age of AI with Pascal Jauffret, Karina Litvack, Brad Harris, and Marya besharov
 
Here is the transcript of the session followed by the Q&A

Marya: Welcome to our final discussion, our final session of the day. It's been a really rich set of conversations and wide range. And what we're going to do in this last fireside chat—and it does feel quite warm actually, we got the fireplace and we’re all going to try to zoom in a little bit. We've touched on so many aspects of purpose throughout the day and what we want to do here is really zoom in on purpose and particularly leadership and the role of leadership and purpose, but with the lens of AI.
Obviously, AI has come up continually; it's already come up in many ways today. But we're going to think about what purpose-based leadership and purpose-based governance looks like in the age of AI. I'm really thrilled to be moderating this conversation for this last session. I'm going to just briefly introduce our speakers and then get them to share with you their insights and experience.
Pascal is the CEO of Forvis Mazars Group which is, as some of you will well know, a leading audit, tax advisory and consulting business. Pascal has had a 30-year career at Forvis Mazars, starting as an auditor, working on the IFRS team, working in Asia-Pacific, then back to Europe, and now as CEO for the past few years. I did know you also have a PhD in strategy and talent management, so we might bring that into the conversation a bit. And you're an Alumni of HEC Paris as well. So welcome, Pascal.
And Karina Litvak. Karina is a non-executive director at Terna, which is Europe's largest independent electricity transmission system operator. She is also the founding chair of Chapter Zero Alliance. For those of you who don't know Chapter Zero Alliance, it supports boards and board members in putting climate and nature really at the heart of corporate strategy and governance. Karina is now an ambassador for that work, so she is still very engaged, and serves on many other corporate boards. We're definitely going to bring in some of this board experience into the conversation because, as we were chatting earlier, governance has come up already a lot, and we want to dig into that. So welcome to both of you.
Pascal, I want to start with you. You lead a firm with a quite strong purpose and a deep history of being purpose-driven in a way that is very much embedded throughout the firm. I would imagine your firm is in an industry that's being rapidly transformed by AI. Talk to us a little bit about how the adoption of AI inside the organization is influencing how you think about purpose and responsibility as a leader.

Pascal: Well, thank you, Maria. Thanks for having me. Can I start by just commenting whenever I see my picture on the screen, it reminds me that I need to put myself on a diet again. No, it is good to be with you and it's an important point you're flagging out. We may comment back on AI and say I'm not quite sure it's transforming our profession as fast or as deep as people are saying it does today. There are a lot of headlines in the Financial Times and other newspapers.
But I'd like to say I think AI sharpens more than dilutes. It sharpens responsibility and it sharpens leadership. It's not a matter of whether AI is impacting our business, it is when and to what extent. And we've taken a very simple thing: we're in a people business. We may come back to it, but we strongly believe that reputation precedes quality, and reputation is a matter of trust and personal interaction. As a result of which, we made the choice to say that AI has to augment our people.
We like numbers, we're accountants, so we tried to measure everything. Today, I would say in the last 90 days, we've identified that AI augmented or freed up the equivalent of 600 FTEs. That's a lot. You can make a big article and newspapers with that: "600 FTEs." But no one got fired. When we nail it down to what does that mean? We have 20,000 people using AI on a daily basis. To make 600 FTEs, it means that AI has brought to each and every one of them seven minutes of free time. This is nothing. So big headlines, big numbers, but you can make numbers lie. What is important here in the purpose is to say we want to augment people with AI, not replace them.

Marya: Thank you, and we're definitely going to come back to what that means for strategy, but also for the individual professionals and how they think about the nature of their work and what augmentation means. Karina, I want to turn to you and ask you about your reflections on this issue from a board perspective. We heard from Pascal about the importance of thinking about augmentation and being cautious about how much it's really changing and how quickly. You're sitting on a bunch of boards, and you've got a really long history of thinking about the role of the board and governance. What do boards need to be mindful of and what kinds of structures do they need to put in place to ensure responsible and purpose-driven leadership in the age of AI?

Karina: Well, let me start by saying that I would never presume to cast myself as the AI expert on the board. I have my doubts about whether—first of all, it's impossible to find enough AI experts to have at least one on every board. So, we all have an obligation to learn about it, to become literate in the same way that climate change is one of those things. I am the ambassador for the Chapter Zero Alliance: it's all about climate literacy on boards.
What I would say here is, when I first started serving on corporate boards 12 years ago, I was the one who said, "Oh, we should have a briefing on AI, please, and cybersecurity," because we need to skill up. Collectively, we really have to master this. And so sure enough, they trot out the Chief Technology Officer. I sat there and I thought: I understand the "and", "because", and "therefore". I know he's speaking English—or Italian, because some of my boards are Italian—but I have no idea what he was saying. Really, none. Zero.
My first thought around this was that we kind of have to develop a common language because we don't have one. In my personal case, it's taken me 12 years. Ten days ago, I had my first presentation from our current CTO and I literally at the end of it said: "This is the best presentation on AI that I have had in 12 years. Thank you, I finally feel like I have understood why you are here and what this means to our business." To which our chairman said, "Coming from her, that means a lot," because I'm pretty critical as board directors go.
I would say that this is a field that is very much left to the "techies" in the firm, and the boards, I suspect, are just being carried along and trying to keep on top of it. That's simply not good enough. To answer your question about purpose, in some ways this is reminiscent of the challenge we face with the climate transition, in the sense that it calls on us to take micro-actions to make our companies better, more resilient, and more competitive. That's what we do as directors.
But it's also the case that we're operating in an ecosystem that is in rapid transformation, and we don't know where it's going. There's one fundamental difference between climate and AI: climate, we pretty much know what we ought to be doing; the scientists have been telling us for 30 years. In the case of AI, even the experts are not sure amongst themselves what should be happening in terms of regulation, oversight, standards, or safeguards. We're starting to feel our way through it. We have broadly three tracks of regulation: the US, which is all about decentralization and innovation; the EU, which is being very cautious and taking a precautionary approach; and in Asia, China is being very muscular, prescriptive, and proactive in pushing for leadership.
When you're a company subject to different regulatory regimes, and when you're a board director, you need to be thinking: how do I even have an impact on an environment I can't control? And yet it has such a huge impact on my company. We need to make sure that we are competitive, seizing the opportunities available to us, and protecting the company, its people, its supply chains, and the environment from the impacts of AI.

Marya: Thank you. We've started at the CEO level and the board role, discussing the complexity—both the technical breakdown and the different regulatory regimes. I want to take us now inside the firm. You talked about augmentation. There's a lot to talk about this: you have a group of amazing professionals who potentially could work at a higher level, but things could also go the other way. What happens to all of that core judgment, decision-making, and situational awareness? What happens to the meaning and purpose of one's work when AI is involved? How are you helping professionals across the firm sustain or evolve a sense of meaning and purpose?

Pascal: Well, very happy too,  but we need to go back in history and take a step back. 30 years ago, when Excel came in, a lot of people said that's the end of accounting; we don't need accountants any longer, we have Excel. Now, people that don't know our profession believe that we have waited for AI to come in to stop keying invoices and processing payroll. For years, all of that has been automated.
AI is not the "big thing," but it does replace some parts. Let's be honest with it, but people do not come to us for tasks. We're not task-based. If I was task-based, I wouldn't hire people from Politecnico di Milano, Oxford, or HEC. I would go for transactional people and have thousands of people in shared service centers, which I don't. Out of 35,000 people that I take responsibility for, people come to me for judgment, and that's the key part.
It may take away some tasks, and probably all of you have been testing ChatGPT, Claude, or Copilot. I suppose you're not taking your Friday afternoon off to go golfing because it saved you so much time that you can work only four and a half days a week. Well, that's the same for us. We have not approached AI as task removal. Of course it does remove tasks, but it's not a skill; it's not like learning a new language. We take it as a muscle. It is something that's going to make you a better version of yourself to serve your clients—maybe faster, with more insights or more color—but you are definitely needed in the mix.
That's crucial for me to emphasize. When it comes to our clients, we don't oppose AI and an agent. We have 3,000 live agents today. I present myself to my clients and say I have 35,000 humans and 3,000 agents. That makes 38,000 "people" that can help you. I use them as a team member. We call it "Bob" because our founder was Robert Mazars, and also because "Bob's your uncle." It's not like, "Wow, we have plenty of stuff." This is here, this is what it's going to do, and it demystifies the whole thing. I think that's what it's all about: you give purpose, you give sense to people, you give direction, and you stop making them freak out about it.

Marya: Thank you. You've described how you've already built this into the way you're thinking. Karina, turning back to you, we talked about professionals and the risk which Pascal has described how they've guarded against. It's a bit more complex when we think about governance. You're sitting as a board member needing to think about accounting for the impact of AI on firms, but also the firm's impact outward. This morning we talked about single and double materiality. Tell us a little bit about how firms are thinking about that double materiality in the context of AI.

Karina: The quick answer to your question is: we're not. We should be, but we're not doing it nearly enough because, quite frankly, right now everybody is so anxious about the so-called AI arms race that we're just rushing headlong into figuring out all the different applications and how we optimize.
In my current board, which is energy transmission, there is such a multiplicity of applications that we are experimenting with and rolling out. By the way, when I joined the board three years ago and I very innocently asked how the whole dispatching system works and if AI was running through the entire thing, our chief engineer said: "We don't use AI at all. This is all engineers who do this." When we had that briefing 10 days ago, it was completely overhauled. It's just one gigantic technology machine. That shows you, in the space of just under three years, how this has transformed.
What we're seeing in the industry is that there was a slow start, but now we're at liftoff. To your point earlier about the impact on people and employment, certainly in Europe we are facing a generalized skills shortage. It is our top risk on our risk register. Right now, it's probably geopolitical risk, but up until not so long ago, it was people. It was getting enough good people because the moment we train people, they get poached; we're a public utility and other sectors pay better. You can get a job in the tech sector for more money than working for the local electricity utility. So, having AI is a way for us to maintain productivity, but we're not cancelling jobs because we're constantly adding jobs right now. That might change over time, but at this point, it's not at all been our experience.
But coming to the double materiality question, this makes me think about the experience from 12 years ago when I entered my previous board, the Italian oil and gas company ENI. That was 2014, and the non-financial reporting directive had just been passed. We had a three-year runway to implement it. We began the task of rolling out that directive, which was then superseded recently by the CSRD, the Corporate Sustainability Reporting Directive. What did that do? It vastly professionalized the practice of sustainability reporting and forced it to be integrated in a much more systematic and robust way with the financial accounts than ever before.
When the material came to the audit committee—of which I was a member and still am today—we are personally answerable for any material errors and omissions. Suddenly, my fellow members of the audit committee said: "Are this data as reliable as the financial data?" Staff said: "Well, not really, because it's really hard to get people to comply and gather the data. So, we use a lot of extrapolation and interpolation. We do the best we can with what we've got." And the committee said: "Not good enough. This time next year, you had better get us data that are as believable as our financial data because our name is on the online. And by the way, we're on the hook for 50,000 euros under Italian law. I say this because this piece of regulation resulted in the company spending 60 million euros to establish data collection and processing systems that then ran very smoothly from that point forward.
So, when CSRD came along, it was a simple matter of just adjusting and adding a lot more indicators. But then when the Omnibus came along and there was talk of rolling back, we just said, "No way. We've invested in this, we're doing it." I think that is an incredibly illustrative example of how regulation changes mentalities. It also changed mentalities because the audit committee said, "Hold on a minute. These data are showing that we're mediocre in this area, and we aspire to excellence across the board. So, come back next year with excellent numbers." Regarding your question, double materiality hasn't reached us in this area yet, and I think it's going to have to be forced on us.

Marya: It's a great example of the power of regulation to really prompt a move. You can't always control that, but when it happens, you've got no choice. One last question for both of you and then we'll open it up for discussion. I want to come back to "AI for good." We've talked about the positives and the challenges on the board level. I want to ask you to look forward: what would a real proactive "AI for good" look like in your firm?

Pascal: I don't think we're looking at anything linear or incremental. Regarding CSRD, the Omnibus is damaging a lot; we can come back to it if you want. But the more I discuss with our clients, they say it's good for them, but it has no material impact on their share price or valuation. This might be a topic for next year: "Purpose and Valuation."
I spent 60 million and I have no questions from the investors about what we're doing on this. I've trained thousands of auditors to provide assurance, yet it doesn't change. I told you I could have tried to impress you with big numbers, saying we are prompting six million every three months on AI. It's a lot, and yet it saves seven minutes on average per person, which is nothing. People ask if we are laying off people. For seven minutes, obviously not.
It's not because we don't see a measurable tangible effect that it is not a good thing. We need to get into it like for CSRD; you don't see the impact on the share price immediately. Luckily, many employees have not said, "Let's reduce costs to boost valuation." We will not lay off people. We will continue to invest in people because that's what I do: I sell people. We are embedding AI into our governance, making sure that a human is always in the loop. Why I'm excited to be here today is that we're not talking about AI per se, we're talking about purpose.
That's what we do as leaders: we give purpose. The one good thing about AI is that it forces leaders to talk about humans. I hate people saying, "AI will transform my business, I'm going to make savings, more margin, and lay off people." What's that value proposition? You could do that 10 years ago by sending work to India; that was a form of "human non-intelligence" at the time. Don't do it now. Invest in people, make quality, make trust.
I love ESG, but I'm not a "heavy E"—I don't use a lot of water or electricity. But I'm a very big "S" (Social). Nobody wants to talk about the "S" because it's hardly measurable compared to metric tons of CO2. I train between 5,000 to 7,000 people every year. That is a big responsibility. If I start my business case by saying I'm going to invest in AI and reduce that 7,000 back to 2,000, I'm not being intelligent. I don't want to do that. We give extra missiles with AI and we'll recruit ten more.

Marya: Your comments reveal why this discussion about AI is so different in this context. Purpose is so deeply embedded in the firm that you wouldn't see it in another way.

Pascal: Purpose is a lighthouse. Nobody needs a lighthouse in daylight; it's completely useless. You need it when it gets stormy and dark. Today, everybody is super anxious. Boards are anxious because no one knows where they’re going to go. Well, then stick to your principles. It's much more convincing than trying to throw numbers in the open.

Marya: Thank you. I'm mindful of the time, so I'm going to open it up to questions. We'll start with Brad.

Brad Harris: Full disclosure, I have a self-interest here. I'm part of a team with the SNO Purpose Center that's putting together a lab to study the future of leadership. One of our big topics is what leadership looks like in the age of AI. Pascal, you said we want AI that sharpens our leaders. I want to ask the other side: how do we know when it starts to dull leadership? Can you give us some rules of thumb for when we stop augmenting and start blunting leadership?

Pascal: That's a really complicated question. In interviews or when facing your board be yourself. I would lie if I said I didn't consult AI before joining to ask for suggestions. Half of the answer seemed convincing because our AI is connected to our own system and all the documents I've been feeding the beast. It was like a super PMO for me, telling me the five things not to forget. But there is no avatar of me today; it's me, and you can sense it in the tone.
Everyone tries to convince people they are ahead of the curve, throwing billions at technology. I've been asking my clients in banking and tech, and they say nobody asks about ROI today, but the ROI is big time negative right now. Yet we continue pouring money in to find best use cases. How could I say the ROI is positive when it only saves my people seven minutes a prompt? If you are true accountants, you might say we are not measuring it right. But even if we're wrong by 100%, it's still only 14 minutes.

Questioner 2: Thank you for your comments. I would love to ask you to speak more about the "S" in ESG. I don't think we need a binary between ESG and purpose. In accounting and reporting, the "S" is incredibly urgent but also messy and noisy. With AI, we have this opportunity to re-understand what it means to be human and human flourishing. I'd love to get your reflections on that.

Karina: Regarding human flourishing: one of my previous board CEOs is an extremely magnetic personality who embodies the values and the mission of the company. People will go to the ends of the earth for him because they are inspired. The risk is that he retires and goodness knows who follows. He genuinely lives and breathes the purpose, and that cascades down to his management.
Conversely, when you have a CEO who operates like an AI model, you can type up all you want about purpose and values, but people are going to become increasingly cynical. From the boardroom, our job is to favor the first type, steer away from the second, and bring authenticity to that mission.

Pascal: Let me complement that with our firm’s purpose. We are in business to contribute to the foundations of a fair and prosperous world, by caring for our people, our clients, and the integrity of our profession. We're not just about profit or growth. We've been training between 5,000 to 7,000 people a year, giving back to the community. Who cares about it? We're not making headlines by saying we give back to the community in Lagos, Nigeria. And yet we truly believe that it fuels our culture like nothing else. I can't give a value or a number to that, but I believe it makes a big difference for people who want to join us.

Paola (Globe Ethics): It really resonated with me what you said about bridging the skills gap and trust being a big element in AI adoption. Karina, it took 12 years to get AI literacy. What elements are important for CEOs and board members to understand about AI now that it is a social-technical topic impacting society?

Karina: First, I don't want to imply that for those 12 years the technical teams were not doing a good job; they were just doing a terrible job of explaining it to us. I was at fault for not investing the time into educating myself. I relied on the company to educate me, and that was a bad decision. I should have realized they were not able to teach me this stuff. Now that I've started to teach myself, I feel I can ask the right questions in the boardroom.
Every director needs to be in a position to say, "Walk me through all the applications you are using AI for," from health and safety to predictive analytics. My boards have been passive, basically saying, "Management's got this." That's the other big issue: the less we know, the less we think we'll be blamed, which is nonsense. Our responsibility is to ask.
From now on, I'm going to be asking a lot more. I pity the next board I join. I want to know: what are the downsides? What are the unintended consequences? Board members need a healthy dose of skepticism about the promises of AI. I believe it can deliver much, but not necessarily everything the "tech bros" are telling us. We also need to think about what constructive role we can play in influencing the context at a national or EU level through trade associations. Even technology companies are worried about this getting out of hand; they want some degree of oversight and safeguards.

Pascal: To complement that: for years, professional service firms were the home of knowledge. AI is a game changer because it disseminates knowledge; you can Google almost any regulation. What doesn't change is judgment. That is why I say it's a muscle. People believe that because you're a CEO, you know best. I am totally ignorant. So I flipped it around: I need 35,000 people to be equipped. Through AI, the level of demand has been elevated. It's intellectually more demanding to be in our profession now because clients are much more informed. My job is to help our people be better versions of themselves.

Marya: Thank you both so much for a really rich discussion. Thank you all for engaging. And it's over to Rudy.

Rodolphe Durand: I'd like to thank the two panelists for concluding our day. It is now almost time for the cocktail. But before that, I'd like us to thank the organizers: Roman Bria, Laurian from communication, the nine PhD students, the Oxford and Poly team, and finally Iiris. You have all received emails from Iiris, and without her, this would not be possible.