Name: State of the Boardroom: Global Risk, Crisis Management and Strengthening Resilience
Uploaded: 2026-02-10T18:32:25.573Z
Duration: 1 h 3 min 4 s
Description: State of the Boardroom: Global Risk, Crisis Management and Strengthening Resilience

Transcript for "State of the Boardroom: Global Risk, Crisis Management and Strengthening Resilience": the NYU Program on Corporate Compliance and Enforcement Program on Corporate Compliance and Enforcement, and my esteemed panel of speakers for a conversation on the state of the boardroom, global risk, crisis management, and strengthening resilience. Today, we're gonna unpack the 2026 risk landscape with two lenses that words rely on. First, policy regulation, and then global risk architecture, and of course, the AI that's running through both. So we'd like to have a practical conversation. So everyone joining us today and thank you to everyone that is taking time to join us. I understand there's over a thousand of you. We'd like you to be able to believe and feel like you've gotten some good information that can shape decision making on the board, what to monitor, and how to prepare. So I'm honored to be joined by three really outstanding experts. Libby Cantrill, the managing director and head of public policy at PIMCO, and she will be helping us interpret the policy and regulatory forces that are shaping board decisions. Maha Hosain Aziz, she's a professor at NYU Program on Corporate Compliance and Enforcement and she's also co chair there in AI policy, the digital economist. Her work in the global political risk area will give us great insights into that. And then thirdly, but not least importantly, we have Oliver Patel. Oliver's head of enterprise AI governance at AstraZeneca. He's also an educator on AI governance and literacy. And he, of course, is at the forefront of shaping AI governance and aligning emerging technology with societal impact. So I think together, we're gonna have an amazing, powerful, multidimensional view of the challenges and opportunities that face today's boards. So let's open the conversation with a quick table setting question, one minute each. We'll start with Libby. What's the single most consequential force you see shaping board priorities today? Goodness. Well, it was nice to be with everybody. I I I would say, I mean, I'm biased because I look at, you know, political and policy risk as it relates to to markets and the economy. So I'm probably going to say, not only kind of geopolitical risk, so kinda headline risk, but also, you know, the risk of, policy follow through. Mhmm. Thank you. Maha, Oliver, either? Sure. Yes, thanks so much, Barbara. It's great to be here. I guess if I had to boil it down to one point to highlight at this point, I would say that we have to appreciate that it's time for a mindset shift. We're not gonna suddenly have certainty about the world order in the next few months or in the next few years even. This is an era of heightened global risks and shock events. Mhmm. And the sooner we have that mindset shift, the easier it will be to navigate what's coming. Mhmm. Thank you, Maha. Oliver? Well, completely agree with with the points that have been made already. I think a specific set of issues that is increasingly pertinent is the impact of AI and AgenTek AI technologies on the workforce and on knowledge work in particular. So it raises really critical questions as to what is the role of the knowledge worker in this new world where AgenTek technologies can potentially take on more and more of the work, and how should organizations balance seizing the opportunities and value of this technology whilst also supporting their workforce and wider society to navigate these changes? Excellent. And so as you can see, we have folks coming from very different directions. And so I think that's just going to make the conversation that much richer. Libby, let's go over to the policy side. What's the headline from your vantage point? What issues do you think are going to nominate, excuse me, dominate the narrative and onboard discussions right now and shape their decisions in terms of policy? Yes. So if I had to characterize 2025, I would say that it, in many ways, was peak Trump. We saw a president who had come into office very organized, mobilized, a real sense of urgency with, I think, understanding Washington much better than he had underserved Trump one point o. So if 2025 was p Trump and more unconstrained Trump, the congress was accommodating. Obviously, Republicans control both chambers of congress. You also saw the courts, you know, which are the coequal branch of government with the legislative branch and the executive branch being very slow to catch up to the president. So 2025 was kinda Pete Trump in terms of an unconstrained president. I think you're gonna see 2026 be a more kind of constrained president where you have, you know, more of the traditional, other kind of tools binding the president in terms of what can actually get done. So, again, 2025, in many ways, I think, peak Trump as it relates to kind of board issues by peak uncertainty. And I think 2026, you'll see, again, many of these other natural constraints start to bind the president. That doesn't mean that he's not going to continue to kinda push the the boundaries of executive authority, but I think you will see, again, the congress assert itself more. I think you will also see the courts, you know, assert assert themselves more and likely will against the president in some pretty important cases. Right. Excellent. And we're going to dive deeper into that. So Maha, how would you characterize, especially on the global, the global and U. S. Risk architecture this year? So, yeah, I think there's no question we have moved into a new phase of our world order. I think everyone is talking about the geopolitical shifts. And I, you know, building on some of what our crowdsourced research has revealed, and this is a collaboration between NYU and crowdsourced consultancy, WikiStrat. We've sort of looked deeper in not just into 2026, but going deeper into this decade. More likely what we're seeing is a push away from US hegemony, but multipolarity, but also moving towards polycentric, a more polycentric world order, where you see that in other regions in the world outside of the West, there is a sense that, okay, we don't need to take our cues from what's happening in The US or in other Western countries. Additionally, we've noted that as we get deeper into this decade, it's gonna be even more clear that other types of actors, the middle powers, of course, and that came up at the WEF at the WEF annual meeting with Mark Carney's declaration of. this is the time for middle powers, and that's important. But in our research, we also track the rise of other types of actors. And I think we we're at the stage where we can ignore the role of the tech firm. You could argue at this stage that the tech firm has as much power as certain governments. They certainly have more understanding of the digital space, and global AI policy is is not realistic. A consensus on that, despite all the efforts, whether it's the UN or other, you know, other groups around the world, it's I think this is it's a great time to be in some ways in charge of a tech firm because you have a lot more influence. You have a lot more ability to drive change. So and and, of course, we can't ignore the billionaires. There is I've been talking about this for over ten years, in fact. But the rise of activist billionaires, it's not just Elon Musk. There are many billionaires that are shaping the narrative. And I think that's something to just be part of one's mindset if you're in the boardroom and trying to navigate what's coming, that dispersion of power. Yeah. Which leads to a much more complex environment, isn't it, to navigate? So Oliver, Exactly. let's let's move to you. We would be remiss if we didn't address AI in the context of all of this. And, of course, there, at AstraZeneca, AI governance in in the larger pharma industry is one of the most complex environments. You've got regulatory privacy, safety, so many competing priorities that you've got to manage. Where do you think So in terms of strategic risk in this AI area are are most underestimating the impacts that are coming. just to address the topic more broadly, and then I'll come to your specific question. I think companies like AstraZeneca, but, you know, many other companies in highly technical or science or innovation focused fields, in many ways, the risks of not adopting AI are as great, if not greater, Mhmm. than many of the risks that come from AI adoption. I think we just, working in AI governance in this type of environment, you have to start with the kind of understanding that there are significant risks, potentially really serious risks for the long term commercial success of an organization in such a high-tech or innovation focused industry to not kind of going all in on AI adoption. So I think that you're seeing this interesting balance between the risks of not capitalizing on AI versus the specific regulatory, legal, and operational, and safety risks that can come from AI adoption without the appropriate guardrails. So just taking a step back, at the moment, the global regulatory landscape for AI policy and governance is pretty complex. So we have these three key jurisdictions here. The US, so so President Trump has made it crystal clear since assuming office last year that the federal government does not want to do anything that could potentially impede. the private sector kind of development and growth in the AI industry. So and that's obviously being progressed through executive orders, repealing some of the some of the AI safety related orders that president Biden put in place, but also through various initiatives to try to kind of deter states from enforcing and passing new AI laws in their specific jurisdiction. So the US government or federal government has made it very clear, Our main policy priority is achieving US dominance and maintaining US dominance in AI. We're not gonna regulate to get in the way of that. China, the second key jurisdiction, interestingly, is kind of as focused as The US on winning the AI race, so to speak, and, you know, kind of putting as much resource and capital and focus on achieving sort of AI supremacy. But interestingly, in China, there are a whole host of regulations in place that do actually restrict and curtail what companies can do with respect to how they develop and deploy AI, but those regulations are quite specifically focused on things like content and information, I. E. What type of content should publicly available AI models be trained on, and therefore, what kind of content is going to be consumed and accessed by the public at large through the use of AI. And then finally, the third kind of key jurisdiction, the EU has is taking a very different approach and basically saying, you know, although we believe in the opportunities of AI, we believe that there are significant risks to individual people and to groups of people such as bias in AI augmented decision making or safety risks in context like medical devices and manufacturing. And therefore, we're gonna treat AI like a product and impose kind of a product safety type framework around the development and deployment of AI, and that's what the EU AI Act is. So for companies that operate globally and for boards that have to think about kind of a global footprint, it's essentially navigating three very different. regimes. US, very hands off, no real AI regulation apart from at the state level, but, you know, that's not I don't think it's quite as relevant as the regulations in the EU and China with respect to, like, a multinational corporation's AI activities. Well, and that, I think, takes us right to Maha Hosain Aziz as we rotate over and talk a little bit more about policy and regulatory and the outlook. And I think, Maha Hosain Aziz, maybe you could bring us to the global risk framework and what kind of macro forces do you think boards should be prioritizing? And especially given what Oliver Patel just described, a really complex regulatory environment with, you know, sovereignty, some deregulating, some regulating, some trying to get the balance right. Sure. It's definitely a complex picture. I think, you know, building on what I said earlier, I think A is the mindset shift, right? That we're we don't really have clarity on the geopolitics and the who's in charge. And I think that will eventually bleed into a polycentric dynamic, which could be positive, Mhmm. but it could also mean heightened insecurity because we don't we don't we're not sure who's allied with who and and which which actors are actually influencing change. I think the other broad trend that I think board members should be thinking about, and again, in our research, it's come up for at least a decade where we've seen this steady decline in governance, in broader democratic governance in particular. In my work, I've talked about it being a crisis of political legitimacy since the Arab Spring, in fact, so even more than ten years, close to twenty years, we've seen that most governments have experienced some kind of severe legitimacy crisis. Citizens pushing back, a steady rise in anti government sentiment. And it's it's leading to this larger question, you know, is democracy actually working? Is it the appropriate environment in which your companies can operate? And that is a that is a significant shift that that comes up in our global risk mindset that, okay, how how are we going to best navigate the risks ahead, whether they're AI related or climate related if democracies are suffering. And according to Freedom House, we've just experienced twenty years of democratic decline. And in fact, most of the decline sharper decline is witnessed in the Western context in The US but also in many countries in Europe. In fact, a lot of studies since the pandemic have shown that younger people, whether they're aging millennials or, you know, Gen Z, they don't really care as much about democracy. They would be okay with autocracy according to some studies and polls. Let's see. They're provided with other benefits. So I think it is we are really at a crossroads as as we get deeper into this decade and deal with the the risks of AI and and climate, as I said, are are governments prepared? Are our companies, our board members going to be in a position to lead when we're not sure if our governments can lead on so many of these issues? And that that's a concern, so just to think about. Excellent contribution there. And Libby, I guess circling back to The United States specifically, domestically, what do you see unfolding now in terms of the policy regulatory kind of trends outlook here in The US? And I I guess, you know, are we gonna get a third shutdown? Is something gonna happen by Friday? I I I hear the senate's working on something, and there's some discussion going back and forth. And and also, there's there's a narrow majority now, in in the house, and things are getting a little tighter. And, of course, I think we're gonna see an election, a major election, which we'll get to next in November. But in terms of right now and and looking forward, starting with this shutdown, what do you see for DHS? Well, I think, you know, just to to build on what, you know, Maha and Oliver said, I mean, to some extent, I don't think this, you know, this particular congress or this US government is in any position to do anything that's sort of forward leaning that is, you know, in kind of beyond corners. I mean, they are incredibly, you know, very much in reactive mode. Just the the the discussion around a government shutdown is kind of a microcosm of that or symbolic of that because it does show that this congress, because of those narrow majorities that you talked about, Barbara, Republicans have the most narrow majority in the house they have ever had as in in their sort of modern kind of state of of their party, that they they are really not in any sort of, you know, majority or in any kind of mindset to do anything that's more strategic. So they really are kind of blocking and tackling, so to speak. And so the fact that we are just talking about, yeah, they can't even fund you know, they they they they we had a full government shutdown back in the fall. We most recently had a partial government shutdown for for four days. Now we're looking at a, you know, January 12 shutdown, but it kinda just yeah. I think for the the the broader point for boards is that this debt is emblematic of a government that in many ways, you know, it's not really functioning, and I think partly because of the narrow majorities and partly because of the political environment. Mhmm. What does that mean though for you know, in terms of, you know, what what can we expect from congress over the next nine months or so until the midterm elections? Honestly, not very much. They did the house did just pass a housing bill. Mhmm. I doubt unless you're in a housing related industry, you care about that. We do. We we care about that at PIMCO because we're the largest buyer of agency mortgage backed securities, so we do care about the the housing market and what have you. But, yes, it's pretty it's pretty kinda kinda cats and dogs, pretty incremental bill, nothing, you know, really revolutionary, doesn't really address the fact that, you know, we have, you know, kinda underbuilt residential, housing by, you know, five, six million units since the financial Mhmm. You know, outside of that, maybe we see some progress on the the crypto market structure bill, that's been having fits and starts in the senate. Yeah. We'll see. You have some pretty colliding forces. You have kind of the new crypto digital assets world, and you have kind of the the old school, you know, banking. world, and and and and there's some conflicts there. So I think the bottom line is that, yeah, we really shouldn't expect, very much from this congress. I again, I don't I think just the fact that we're talking about, you know, another shutdown is sort of emblematic of that. The last thing I will say though is the president is going to lean into this affordability, you know, issue. Mhmm. He's apparently told his the the White House not to talk about it as an affordability, but more of cost of living because affordability is more of a hoax or what have you. You know, say what you what you will about that. But he is going to lean into this. And so, you know, again and if you're, you know, in the markets or even as a board member, you may see a lot of headlines and think, wow. You know, this is this the this president is, gonna do a lot before the midterms. I would just say that the reality is that most of what he has talked about, he needs the congress for. So for instance, you cannot impose a 10% max, you know, interest rate on credit cards without the congress. There is no, you know, there is no law that allows the president to do that unilaterally. He cannot prevent, you know, institutional investors from buying single family homes unilaterally. You need also congress for that. So I would just say that the the rhetoric here, you know, like in many times, is quite different from the reality. And so while he's going to talk a lot about cost of living and some of the things that, you know, he may want to do, again, I think the reality here is is quite different. The last last thing I will say is I think Republicans in on the hill are very much hoping that once these are tax, cuts from the one big beautiful bill that was passed last July start to really work their way through the economy. And you're already seeing this with lower withholding tables, so people have higher take home pay. But you're also going to see this with some of the highest refunds that, you know, folks have have gotten, you know, anywhere between 1,000 to $4,000 per household. Household, obviously depending on kind of your tax liability and your tax situation. Mhmm. But the Republicans are hoping that that will do it, that that will make people feel better, that the sort of k shaped economy that folks were talking about in 2025, that was largely driven by high end consumers and then CapEx, you know, related to to to AI. The AI build out, what have you, will start broadening now, and that voters will feel better about the economy. It is a area that is a real liability for the president. He does not pull well on the economy. And so I think, you know, Republicans who are up in the midterms, of course, he's president Trump is not, but Republicans are, are hoping that you're gonna see a broader, you know, economic gains from that from that tax bill and that people will just voters will be in a better mood than they are right. now. Well, I'm not gonna put you on the spot and ask what you think the outcome would be as a result of all those those steps in November. But but let me stick with with you, and and let's just assume that in November, we do have a truly divided government. What do you think the implications of that would be for, obviously, the next couple years? Would we would we see for both fiscal policy, the markets, the economy? I mean, are we gonna see more executive orders than you know, we're gonna see another like, as you said, year one of Trump, this this particular term, or something completely different. And and or do you think there are areas of potential compromise that would be heartwarming for corporate boards? Yeah. Well well, you know, we're investors, and so our investment committee is always putting me on the spot to make predictions. So I'm I'm happy to I'm happy to do that. I think that just, you know, passes prologue, the house will very likely flip to democratic control. And just remember that, you know, Republicans, as we talked about earlier, only have a three seat majority in the house. And the average seat gain of the party out of power, usually the midterms are sort of a referendum on the current part part, you know, power current party in power, is 26 seats. So Democrats really need to have just, a so so night in order to actually take back the house. Now Democrats, we could talk about that for an entire session. They're kind of in the wilderness, and then, you know, they can get in their own way as well. But, again, they only have to have, like, a mediocre night in order to take back the house. The real question, I think, is the senate. The senate right now, Republicans control the senate by four seats. Usually, the party out of power actually wins on average four seats. So you would sort of think that that means that the senate is in play. Yeah. There is a sort of expression geography is destiny in the senate. In particular, the senate seats that are up are very difficult for Democrats to to to not only defend all the seats that they they all have up in terms of Michigan and and Georgia, both of those states president Trump won, and they're defending Democratic senate seat seats in in those states, but also, you know, winning four of the five that are sort of in play for Republicans, that would include Iowa and Ohio and Texas and Alaska, they'd have to have a very, very, very good night, basically, a perfect night in order to take back the senate. So kind of a a mediocre night to take back the house, but a perfect night to take back the senate. So I think that the base case, as you referred to, is probably split government in 2027. That's when the new congress would come in to into power. And I think they're pretty there are three big implications. One is you basically don't see any real legislation. Now I think it may be a different political environment. There may have been some opportunity for compromise. Think immigration. If you get, you know, Republicans and Democrats in the room privately, I think they would admit that our immigration system is broken and that we need legislation in order to move forward. Maybe you could see something in AI and what have you, but I don't think that the political environment. going into what looks like that kind of a contested general election in in 2028 will sort of lend itself to, you know, big productive bipartisan legislation or negotiations. I could be wrong, but I I think it's unlikely. So I think the base case is, you know, probably very little passing congress, sort of a stalemate, if you will. Democrats really trying to focus on, like, symbol symbolic bills, but not really anything get actually gets passed both the house and the senate signed into law. So that's one implication. Second implication is that just remember that in the house, not only if Democrats take back the house, they'll have control of the legislative calendar and what's voted on and the committees and what's brought up in those committees, but they also have subpoena power and subpoena authority. And that means that you're gonna see a lot of kind of oversight from a democratic house, kind of throwing sand in the gears, if you will, of this administration, and that will slow them down. So even though you said that you you might see them push forward with executive orders, they're going to be, not necessarily derailed, but really slowed in terms of what they can actually push forward because they're gonna be spending time on, you know, getting getting documents together for the for the hill and preparing for testimony and what and what have you. And the third implication is that even if Democrats don't take back the senate, and again, I think that's that's quite unlikely, it is, you know, most likely that they the Republicans have a more narrow majority in the senate than they do today. Right now, Republicans have a 53, 47 majority. Why does that. matter? Because for most legislation, you use 60 votes. But for to confirm a presidential nominee, you only need 50 votes. Mhmm. And so that practically means that if they have a smaller majority in the senate, then president Trump is going to again be more stymied or or or sort of slowed down in terms of who he can actually get into these positions of power. You know, that expression washing personnel is policy. And so president Trump can't put exactly who he wants at the Fed, for instance, then, you know, that will have sort of implications. So, again, you know, not legislation of any, subpoena authority which can throw sort of sand in the gears, and then again a harder time for president Trump to get in the personnel that he wants because of what's presumed to be a smaller majority in the senate. And so implicit in what you've said, think, is that the gerrymandering we see going on in various of the states and then the counter gerrymandering is not gonna have a material impact on the outcome of the election in November. Got it. Alright. No. It's kind of I think it's gonna sort of negated. The only thing I will just sort of, you know, put in people's on people's radar is that the supreme court is likely to issue a ruling on the Voting Rights Act, and that could actually have some pretty significant implications for the house. Not in 2026 because a lot of those districts have already been drawn, but for 2028 and thereafter. And depending on how the supreme court rules, it basically could make it much harder for, you know, state legislatures to draw, you know, majority minority districts. And that likely go in favor of Republicans. That means that Republicans would have a structural advantage in the house. But, again, that's not going to be for 2026, but it could be, you know, a phenomenon in 2028 and then beyond. Yeah. And then we've got the 2030 census. Right? It's gonna also be a factor out in the future. Yeah. Let us move. Yeah. I mean, the the yeah. The census seems to be less important now for redistricting given all these states are just choosing to redistrict whenever they want to. I mean, California and Texas have so the census seems be super important. I think it's actually less important now. Yeah. Yeah. I I think just on the size of the delegations relative to, you know, the movement. Yeah. Yep. Oh, and and, obviously, for the general elections, Yes. for the electoral college, it is still very important. So, Oliver, I'm sorry to leave you out of the discussion too long. Let me get over to you. The EU AI Act has stringent obligations for high risk AI systems. What aspect of that act do you believe the boards are underprepared for, especially around data governments governance and documentation burdens? Well, I was thinking as you were having the discussion that I don't mind making predictions about the future, but I was also given really good advice earlier on in my career when going on TV interviews and on panels, not to stray too far into topics I'm not comfortable with. So I I'm glad you're not asking me to predict what's gonna happen in the midterms is all I'm saying. because. it's not my area of ex but I can I can talk about the EU AI act and also more broadly about the EU's kind of shifting the shifting sands of EU tech and digital policy and regulation more broadly, and then the EU AI Act specifically? Because what's happened in the past few months is that there's been a really robust debate within the EU as to whether the kind of digital rule book that the EU has established over the past few years to regulate AI, to regulate online platforms such as social media platforms, to regulate data and personal data and information. There's been a really robust debate within the EU as to whether that whole rule book and that whole framework is essentially too complex and making life too difficult for for business. Mhmm. And so, actually, what happened is, I think it was in in 2024 or perhaps early twenty twenty five, Mario Draghi, the former president of the European Central Bank, published a report that is known as the Draghi Report that basically made a really compelling argument that the EU is having really significant challenges in terms of productivity, in terms of economic growth that could really have long term detrimental impact on the EU's competitiveness. And there was quite a lot of attention paid in the drag year report as to how the impact or the potential long term impact of overly complex and perhaps difficult to comply with regulations and the ability of digital startups and scale ups in particular to really compete internationally. And this report has had a massive impact on what the European Commission is currently doing. So even though it was kind of, I guess, independent report, you could say, in a sense, it wasn't the formal position of the European Commission. It's had a big impact on what the European Commission is doing now. So back in November, the European Commission published what's known as the digital package, which was basically a set of proposals that are intended to kind of simplify and streamline this set of laws that has been passed in the last few years, with the EU AI Act being one of them. So I will talk a bit about the EU AI Act in detail to address your specific question. But I just wanted to give the context that right now, there's a really big debate within the EU as to whether actually simplifications and changes need to be made to all of these laws that actually having a really, like, global impact. And this kind of concept of the Brussels Effect has become quite. pertinent in the last few years because the tech laws that the EU passes and the digital laws and standards that the EU passes essentially, in many ways, become adopted by companies and multinational organizations, including US tech companies, obviously, who want to do business with and and provide their products and services in the EU. And what we've seen interestingly in in in 2025 in particular was quite strong pressure from from different parts, maybe different parts of the US government and US industry on the EU about this about this very set of issues. So whether it's GDPR, EU AI Act, Digital Services Act, the kind of core tech. that have been passed in recent years, boards need to be aware of the fact that, actually, there could be quite meaningful amendments and changes to these laws, which are and these amendments are primarily designed to essentially make life easier for companies and improve growth and productivity in the EU. So on the EU AI act specifically, it's it's an interesting situation. Right? Because the EU AI act entered into force in August 2024, and the first set of provisions became applicable in February 2025. And the proposal to amend it was published in November 2025, Mhmm. which I think highlights that it's not easy to regulate AI. It's it's moving so fast. The tech is changing so fast. Our understanding of risk is changing so fast. The word agentic, an agent, isn't in the EU AI Act. So none of this is easy, right, keeping up with this technology and governing it. I think in terms of what boards and companies more broadly are sort of I don't want to kind of accuse people of being overprepared or underprepared, but it's more I think actual compliance with the EU AI Act requires a level of oversight over the full set of AI activities within the organization and a kind of system in place to actually keep track of, whether it's through an AI catalog or an AI inventory, to keep track of all the different use cases and AI systems that are in production and in use, it's really difficult to comply with laws like the EUAX without that kind of three sixty degree view of what your organization is doing with AI, which systems are in use and what are they being used for, which data is being used. The challenge is AI is democratized, so it's very difficult keep track in a kind of centralized way. to give the board and the governance functions that three sixty degree oversight they need. It's very difficult to actually manage that in an environment where anyone with a phone or laptop with Internet access can spin up an AI instance or use case using publicly available technology. So the fundamental tension there is how do you design for compliance with laws like the AI act, which require quite broad kind of AI inventory and and that kind of oversight. So you can actually know what's going on so you can intervene where where stricter governance is required versus enabling and empowering your workforce to use AI and use as much AI as possible to to make their kind of, you know, efficiency, productivity, and and and other kind of transformative benefits. So as with everything in AI governance and policy, it's that kind of essential balancing act between managing risk, complying with regulations, and maximizing the value of AI. And a part of that now is about democratizing access to AI. But, of course, the more you democratize, the more you make it easier for people to use AI, the harder it could be to control and have oversight over everything, which is obviously a core part of governance. Okay. So given this in Europe, when we when we think about cross border data flows and transfer constraints that are evolving, What disruptions should multinational boards anticipate? It's ironic to me that we're talking about trying to smooth very complex cross border data flows when this is something I remember at ITT when I worked there thirty years ago. I would go over to the UN to try to figure out what can the UN do to help with transborder data flows. So thinking now about a a board that's a multinational board, what are you suggesting that they do? And in terms of trying to get ahead of, and I guess if it's impossible it is, these disruptive trends that are so complex to get your arms around. Yes. It's absolutely great, great point. So some practical solutions. So first of all, actually, before I worked on AI governance at AstraZeneca, actually worked for the UK government, and I was part of their international data flows team. And my role for the UK government was to kind of support and lead the work on securing the data flow and data sharing agreement between The UK and the EU. as The UK exited the EU. So it's a topic I'm very familiar with. I think when it comes to how AI technologies are now democratized and the challenges of oversight, I think there are some The key point, right, and it's maybe a bit counterintuitive, is actually the biggest risk is the kind of workforce, is shadow AI, and the workforce just using whatever AI they can get their hands on that's publicly available. Because once that happens, you don't have any real control over what people are doing, where they're sending data, what services they're accessing, and what they're using those services for. So really, the companies that are kind of at the frontier and really forward thinking, they understand that one of the best things you can do to actually manage risk is provide your workforce with access to best in class AI capabilities internally, a, so that you can get the kind of benefits of doing that from, like, a productivity and innovation perspective, but b, you actually significantly reduce the incentive for anyone to leave the kind of enterprise firewall when it comes to their use of AI. So you provide access to best in class AI capabilities internally. You make it very clear which of those capabilities are approved. Mhmm. So I have a kind of three g's framework for for dealing with this. So you've got green light, your approved AI capabilities. And the trick there is to continue making sure that you're actually keeping up with what's at the frontier, which changes which changes all the time. The second g is guardrail. So once you have your approved capabilities, you figure out what technical guardrails and controls need to be built into them to address concerns like the EU AI act or like data sharing. Mhmm. And then the third g is guidance. Because in an era where everyone can use AI, you're not gonna have manual governance and risk assessment of every use of AI. It doesn't work like that. So really, the the training and the guidance you you provide to the workforce is then critical. So. AI is democratized. As you've alluded to, it increases compliance risk, it increases cross border data flow risk, among other factors. But through following those three Gs of green light, guardrail, and guidance, you have a bit more of a chance to actually protect your organization whilst also enabling the organization to innovate. Excellent. Oliver, thank you for that. And I before we leave disruptive forces and really focus in more deeply on AI, I don't wanna not circle back to to Libby. There's so many disruptive forces, midterms, tariffs, trade, the US dollar, interesting things there. A lot of folks buying gold, including sovereign nations. You the the monetary policy and what's going on with the Fed. So, Libby, I'm gonna turn that over to you and and maybe you can, in in some of the time we have remaining, give a focus on where you think the biggest disruptions are that would be having board impacts, not only over the next year, but maybe the next three. Yeah. Not a super easy question to answer, but I may let let me just sort of tick through the three that you you mentioned because I think those are, at least from our client base, some things that are kind of front and center. You know, in terms of tariff policy, you know, this this president has been very clear. He loves tariffs. We should take him at his word. He really does believe that tariffs are both a means to an end to increase his negotiating power, what have you, but also an end unto themselves. And so don't underestimate that part. Like, he actually is fine with tariffs as a destination. With that said, and again, sort of consistent with this theme of we're past peak Trump, I do think that the Supreme Court will probably rule against this administration on what is called IEPA, this International Emergency Economic Powers Act. That's a statute that the president has used for about half of the tariffs. Right now, you know, we we saw the tariff rate go from about 2% in January 2025 to around sort of 13% right now. The about half of those tariffs have been imposed under this IEPA statute. The the supreme court, as early as February 20, next Friday, may rule against the president. However, just keep in mind, the Trump administration has a lot of other tools, where they can impose, that they can impose tariffs on. So I think that, you know, tariff policy uncertainty again, inconsistent with this idea of peak Trump, that we're kinda probably past peak tariff policy uncertainty and probably past peak tariff. That was obviously on liberation day of last year. But I do think that we're gonna have to live with tariffs, not only for the rest of the Trump administration, but probably thereafter. I mean, Democrats, the sort of dirty little secret is actually Democrats were the original protectionist on Capitol Hill. So I think you're gonna have to kinda live with that uncertainty, although we're probably past the peak. In terms of the dollar, and sort of this debasement, dedollarization trade, you know, I think our view is that two things can be true. One is that the dollar continues to be the reserve currency. It's just dollar is still dominant. The dollar is still king, in terms of, you know, the transactions being settled in the dollar, commodities being priced in the dollar, payment system, what have you. With that said, you can also see the dollar reprice. And, again, this administration, maybe they want a little bit of a weaker dollar that helps exports and what have you. But we also think that, you know, the the the economic conditions have also changed. And as a result, you have seen the dollar went from its real high, right, coming 01/20/2025, you know, reprice. And we think, again, those those both of those things can be true. In terms of gold and the spirit of other, like, dollar proxies, you know, I think our view is a lot of this is retail. It's not institutional. You actually look at central bank holdings of gold. They really have not changed if you actually keep the price of gold stable. So what it looks like, it's a stock effect, not a flow effect, meaning that you actually haven't seen central banks really increasing their gold allocation. It's just that the gold that they had on their balance sheets have been increased because the value the prices increased. So it's more of a market value effect, again, than a flow effect. So I think we would feed this idea a little bit about gold. That said, we do you know, from our from our perspective, lots of, you know, good opportunities outside of The US, whether it's in developed markets or emerging market, you know, emerging markets or or what or what have you. And then lastly, in terms of the Fed, you know, I think my own view, I've been pretty sanguine about Fed independence. I haven't really been worried about it. Lots of guardrails to keep the Fed protected from political interference. Mhmm. You know, the fact that the chair is vote, the fact that it's actually a 12 FOMC committee, not just a seven fed board governor, the fact that there is this senate confirmation process, and I think we have seen that play out. You know, the president probably wanted Kevin Hassett, his NEC director, someone who he thought was going to be more pliable, who would be cutting rates and what have you. He didn't get him because the senate, you know, basically told him that he would be that he could get confirmed. We have Kevin Worsch, my actually, my former vice president when I worked at Morgan Stanley at the beginning of my career, somebody who has talked a lot about Fed independence. Mhmm. You know, he well, he'd be a bit more dovish than his, you know, previous Fed speeches when he was, you know, at the Fed many years ago, probably. But, you know, he he has some strong ideological views about, you know, keeping the balance sheet small, and, again, about, making sure that the Fed is independent. So I think our view, my view, personally, had always been, you know, more sanguine about this, you know, concern around Fed independence because of lots of guardrails. And I think that's, you know, punctuated with this election of Kevin Worsch. And I think we we expect him to, you know, make those monetary policy decisions based on the the the data, not necessarily based on the politics. Mhmm. Okay. And, do you have a forecast on do you think he'll the Tillis situation will go away and we'll have him in place by the end of Jay's term just as chair, not as governor mid May? Yes. So Powell's term ends in in in mid May as as chair, as governor. It doesn't end until January 2028. Right. I, you know, I think kinda TBP. As your senator Tillis, you know, you're he is the swing vote. He is the one who basically unlocks the confirmation process. I don't think Kevin Worsch will have any problem getting confirmed once that process but he's also said that he will not start that until, the the president withdraws the criminal investigation into Powell. I I I I was surprised. I thought that Powell that that Trump would try to walk back from that investigation. He instead, he's doubled down. So I I do think that is the thing to focus on. What the markets are not necessarily pricing, though, is for a Powell fed you know, Powell to continue as chair, so de facto chair beyond May, because, you know, that investigation is still ongoing. Yes. Trump has said that he could continue that until Tillis steps down in January 2027. So, again, I it's not necessarily a base case, but it is definitely a scenario and probably not too unlikely of a scenario that Powell stays on for, you know, longer than what the market is pricing in. We do think though the economic data will likely support at least one rate cut, if not two. Inflation is coming down as tariff pass through, you know, decreases. And if the economy is is somewhat resilient, but the labor market is a little bit weaker. And so for those governors who are focusing on that kind of weakness, we do think it would support, you know, again, at least one rate cut, if not two, at the second half of this year. Great. Thank you, Libby. I'm going to jump to my high. I know you've been wondering, wait a minute, my turn. And it is your turn. So on AI and emerging tech cutting across geographic, economic, societal dimensions, where do you see AI amplifying or accelerating the broader risk categories? And where do you think boards today are underestimating AI's role as a destabilizer for their businesses? Or on the opposite end, missing a real opportunity, which I think I. referred to earlier. It cuts both ways. Mhmm. Sure. No. In our in our work, again, we in our CrowdTest research, we've we've looked at global risk, but AI obviously has come up quite a lot in the last two years. I think most recently, we've been noting this backlash against data centers. globally, where you have citizen movements that are, highlighting through protests the fact that, obviously, there's an energy problem, environmental imprint, associated with these data centers. So I I think that's something to keep in mind. It's easy to just overlook it, but that trend is growing in this era of heightened civic engagement. I think that's protest trend that will will accelerate. And as part of this broader tech lash, we've also noted well, I mean, I think everyone has talked about AI, the impact of AI on unemployment. I think and there was a great study done by Stanford's digital digital lab that pointed out that already you see that early career roles are being just are not being offered. So people coming out of college. at the moment are not I mean, I don't know how they develop their career if they don't have a starting point. So I think in our in our research, we've really come up with this idea of an occupational identity crisis that I think most governments and even some companies are overlooking. What happens to our societies if you have more people who are not just unemployed temporarily, but have fallen through the cracks in terms of career progression because AI has replaced that early phase? And I think that could potentially bleed into mental health crisis. I mean, if you think about our purpose and our work is such a big part of who we are. And if that's taken away, I think that will bleed into bleed into, social instability that I think most people are not taking into account. I think companies should go slow in terms of AI adoption to see to what extent they can, upskill or reskill people who are already there as opposed we talked about this at the New York AI Summit in December. Should you be the tortoise or the hare in terms of AI adoption? And I'm I'm definitely for the the slower approach because otherwise there will be significant backlash and, again, a broader techlash at the level of jobs, at the level of data centers. And and I think it's only going to given we have a problem with regulation and as Oliver said earlier, how do we regulate something that is evolving? I think this is the the AI risks and the impact on society and economy cannot be ignored. Just last note, I I was just reading recently. I'm sure you guys have heard about MaltBook, which is a social media platform just for AI agents that is independent of any any human input is talking about suing humans. And, you know, I mean, I I think we should be alarmed. I mean, all the discussions we're having, how does that shift when we have AGI, when AI doesn't really need our input? I don't know what that means for the job market. I I don't know if I believe Elon Musk about the universal high income that he's promising. us because do do everything for us. So I think much of our discussion is sort of near term. I think if we think longer term, that transition to AGI, I think that should worry us as well. Yeah. Good point, Maha. That took my breath away when I read about that yesterday. Yeah. Just accelerating so fast. I think, Oliver, here's a question just circling to you. Where do you see given all of this rapid innovation and and, of course, guardrails kinda lagging behind, if you will, at least in this country, Where do you think boards are overestimating AI's capabilities or maybe underestimating? Great question. So I think that what's happening now is people are starting to really understand just quite how capable these new AgenScik AI systems could become. So there was a study done by the UK AI Security Institute on Frontier AI Trends published at the 2025, which basically found that for software engineering tasks, about 40% over 40% of the time, AI can successfully complete a task that would take a human software engineer over an hour. So we're talking about specialists who have trained for years that could spend an hour on you know, at their terminal or their computer working, and AI agents can already do that nearly half of the time. So you can just extrapolate into, like, the future just a little bit. One year, two years, three years, that's gonna rapidly go up, and the the the time kind of human task completion time rates and what AI agents will be able to do will probably increase as well. But where there might then be a slight misunderstanding is that, oh, well, we just don't need the humans anymore. The AI agents will do it. But that's wrong because all of the research in this space, and honestly, of the anecdotal evidence that you just see in working across industry on this, is that the best results come from pairing AI capabilities with subject subject matter experts, people who have deep expertise on a business process or on a problem or in a specific area, but also have the AI literacy to understand both the limitations of AI and the capabilities of AI. So it's like you you kind of have this holy trinity of deep subject matter experts in a specific domain, technical experts that can help them build something that's gonna work, and the advanced tech itself is becoming increasingly capable and autonomous. Now, the challenge there is what about all of the, you know, people at an earlier stage of their career, the graduates and others, that are not yet deep subject matter experts? What do they do? Well, I don't think the answer for organizations should be we're not gonna hire them. Because ultimately, all of the all of today's deep subject matter experts are one day gonna retire. So there has to be some kind of long term thinking here where it's okay. Maybe and to be honest, I use AI loads for my for my work. And to be honest, there are various ways in which you can you can think, wow. A few years ago, I probably would have hired someone to do that, and now AI can do it. And so I can understand the temptation to think just, you know, we will have way fewer people. But what what's the long term play with that? Because ultimately, you'll you need people who actually understand how to get the best out of AI, how to build applications that are based on human expertise and human context, which AI is not very good at picking up unless it's really optimized in that kind of way. I'd be remiss if I didn't mention the fact that coming out in September, I've got a book covering all of this called Fundamentals of AI Governance, and you can find it at aigovernancebook.com where I'll be exploring how all of this impacts enterprises and just the broader landscape. Well, thank you, Oliver. And I think these trade offs that you're describing are so important for boards to navigate. What about an adaptive leadership model for boards? How do they build one? You know, how do they address what was just described there by Oliver? I think you're on mute. Sorry. Yeah. Yeah, that's a good question. I think, you know, I guess I'm more focused on geopolitics. Would say that board should have an in house chief geopolitical officer to navigate what's changing. I mean, you can't have you can have an opinion today and tomorrow. It could be very different because just of the way the world is shifting. I also think with respect to AI that I think from what I've noted is that chief legal officers, because having engaged with them through World Economic Forum and other. settings as well, that they seem to be in tune with the effect of AI on unemployment. I think it's not I think that's a useful strategy. I don't know if that has been part of the role in the past, but they've definitely adapted. So I think there's I think there are ways to to make sure working with HR, working with your chief legal officer to see what is the best strategy so that we don't experience a backlash, a techlash, Okay. so that we can prepare our workforce and keep, you know, keep this make sure this is a human driven transition, a human centered transition. Because sometimes when you hear the technologist, it gets it's really easy to get carried away and just think, okay, AI is the best thing and we we don't need to worry about how this impacts the individual. Okay. Alright. Can I just I just checked it with one one just just one kinda overlay to this idea of this tech lash? I mean, we saw this in off cycle elections in Georgia and Virginia. November 2025, this idea of, like, data centers and just the affordability issues being exacerbated by increases in energy prices and what have you. And I think that this could be an issue in the midterms, so I think even more likely, this is gonna be a general election issue in 2028. And so this is something you just think, you know, that the Europeans, you know, maybe regulate rather than innovate. Sorry. That might be kind of harsh on our European friends. But but here in The United States, I do think that there is going to be a potential political backlash, to AI. And so what what that means from a policy making perspective, yeah, I think it's an open question. But I do think that the politics of this, not just in terms of labor displacement that Maha has has talked about and I think understandably worries about, but more just in terms of exacerbating some of these already these affordability and cost of living issues that are already. kind of, you know, forefront. and paramount in voters mind. very good point, Libby. I think I'm sorry to say we have about a minute and a half left. But I think what we'd like to do is, thirty seconds each, if you could share with us, if you were gonna do one slide to brief a risk committee, what would be on it? And I don't care if it's an opportunity or a risk after having been through this entire discussions. And why don't we start this time with Oliver? Well, I think that what I would focus on is how to balance the risks of not adopting AI with the risks of adopting AI. And I think the risks of not adopting AI could be around your kind of ability to innovate, ability to to keep up with real competitors and disruptors in your industry, and also to retain the best talent. And the risk of AI will be a lot of the stuff we've spoken about, the regulations, the compliance, the potential backlash, the impact on the workforce. So getting that balance right is like a million dollar question right now or maybe even a billion dollar question. K. Libby? Yeah. I mean, I think just to sort of draw on the things that we talked about, know, political uncertainty is going to be is not going away. So even though, again, I sort of said we think we're past peak Trump, the not only here in The United States where I think political and policy uncertainty are going to continue not only under this administration, but, you know, beyond, you know, 2029 and what have you. If you just think about the combination of of AI, but also demographics, you know, indebtedness. All of these are kind of a, you know, a stew, I think, for, market persistence and and boards, and it's it it is just going to be a question of navigating around that. But then internationally, I mean, this sort of more fragmented, you know, more polarized world, just, I think, means that there probably will be more opportunities by more inefficiencies, honestly. And also, you know, risks will just be the same. It's kinda, you know, the new normal even though that's a, you know, actually, one of the founders of PIMCO actually, you coined that that that expression a long time ago, but this is gonna be the new normal. So just making sure that you do have folks who are covering these geopolitical and political risks because, you know, unfortunately, they are not going, you know, Right. away anytime soon. Excellent. Maha. Mhmm. Yes. I agree with what Oliver and Libby have said. I think this is an as I said earlier, this is an era of heightened global risks. and shocks, and it's not going away. So having the in house chief geopolitical officer or an empowered chief legal officer to navigate this is important. And I think it's also important to reflect on what happened at the start of this decade. I know we've all put the pandemic behind us, but what did companies do, what did boards do when we were dealing with this major systemic shock? I think that playbook is what should be implemented going forward because it's just going to be it's as though we're having, you know, that systemic shock on a regular basis now in geopolitics, politics, economy, society. And again, going back to the idea that this is a we live in a crisis of political legitimacy globally. It's not just one country or one developed country or developing country. It's global. So having that playbook from the pandemic could be a good strategy going forward. Well, thank you all. I think it's been a great conversation, Libby, Maha and Oliver. I think we could have gone on for another hour or two, but I think we have to wrap it up here. And of course, thank you to our hosts again. And everybody, I hope you enjoyed this and have a great rest of the week. Thank you. Thank you. Bye.