Listen to The Bid – BlackRock’s investment podcast

211. The Intersection of AI and Geopolitics From An Investing Lens

Episode Description:

Rapid advancements in artificial intelligence are influencing global power dynamics in ways that are reshaping the investment landscape. So, how are these AI developments impacting geopolitics, and what does this mean for investors?

Catherine Kress, Head of Geopolitical Research and Jeff Shen, co-head of systematic equities at BlackRock join Oscar to help us understand the current themes driving the intersection of AI and geopolitics, how the broader technology and AI race is evolving, and the broader implications of the shifting US policy landscape on active management opportunities.

Sources: DeepSeek model training cost claims from paper, January 2025; 60% of US tech revenues derived from overseas based on Aladdin Explore, using revenue breakdowns from the S&P 500 index as of 4/30/2024; DeepSeek's low-cost AI spotlights billions spent by US tech Reuters, January 2025; How DeepSeek AI Sparks Nearly $1 Trillion U.S. Tech Implosion, Investors Business Daily, January 2025; Nvidia Stock Has Almost Recovered From DeepSeek Rout. There’s More Good News Barrons, February 2025; US government commission pushes Manhattan Project-style AI initiative, Reuters, Nov 2024; US tightens control on AI chips export drawing pushback BBC, January 2025; FACT SHEET: Ensuring U.S. Security and Economic Strength in the Age of Artificial Intelligence White House Press Briefing, January 2025; What Donald Trump’s Win Means For AI Time Magazine, November 2025; Trump plans to impose 25% tariffs on autos, chips and pharmaceuticals, CNN, February 2025; REMARKS BY VICE PRESIDENT VANCE AT THE ARTIFICIAL INTELLIGENCE (AI) ACTION SUMMIT US Embassy 2025, February 2025

Written disclosures in each podcast platform and each episode description:
This content is for informational purposes only and is not an offer or a solicitation. Reliance upon information in this material is at the sole discretion of the listener.
Reference to the names of each company mentioned in this communication is merely for explaining the investment strategy and should not be construed as investment advice or investment recommendation of those companies.

For full disclosures go to Blackrock.com/corporate/compliance/bid-disclosures

<<THEME MUSIC>>

<<TRANSCRIPT>>

Oscar Pulido: Rapid advancements in artificial intelligence are influencing global power dynamics in ways that are reshaping the investment landscape.

Catherine Kress: artificial intelligence or AI has become this driving multidimensional force in the geopolitical landscape. and nowhere is that clearer than in the technology competition between the US and China

Oscar Pulido: So how are these AI developments impacting geopolitics, and what does this mean for investors?

Jeff Shen: I think we're entering a period we are likely to see more reversal, more contrarian movements, it could be a big news for a day and then things start to revert. a, a more of a prevalent behavior in the market.

Oscar Pulido: Welcome to the bid where we break down what's happening in the markets and explore the forces changing the economy and finance. I'm Oscar Pulido.

Joining me are Catherine Kress, head of Geopolitical Research and Jeff Shen, co-head of Systematic Equities at BlackRock. They'll help us understand the current themes driving the intersection of AI and geopolitics, how the technology and AI race is evolving, and the broader implications of the shifting US policy landscape on active management opportunities.

Catherine and Jeff, welcome back to The Bid.

Catherine Kress: Thanks Oscar. It's great to be back in the studio with you.

Jeff Shen: Great to come back, Oscar.

Oscar Pulido: And, I say, welcome back. it's been very recently that we've had both of you on, but we needed both of you back on because we want to talk about two of the mega forces. We want to talk about geopolitics and AI. And these have been topics that both of you have joined us on the podcast to talk about.

Catherine, if I could start with you. Tell us a little bit more about this intersection of geopolitics and AI that is occurring right now and what are the core issues that you're looking at?

Catherine Kress: So, Oscar, you and I have talked about this a number of times on previous episodes, but to set the scene, artificial intelligence or AI has become this driving multidimensional force in the geopolitical landscape. It's at the white-hot center of global Geopolitical competition and nowhere is that clearer than in the technology competition between the US and China. The US and China are engaged in this epic zero-sum, long-term structural competition to seize the commanding heights of advanced technologies in the 21st century.

Now we've seen this technology competition play out across a range of dimensions and touching on a number of different technologies. But what's front and center right now – and the topic of our conversation today – is artificial intelligence. Why is that? First, there's a strong belief both in the US and China, as well as in many other countries in the world, that AI will bring tremendous economic advantage to whichever country that leads in it.

Second, there's also a belief that AI will spark a revolution in military affairs, and we're only at the front end of this. The potential use cases for AI are innumerable and have the potential to radically change everything about how militaries around the world organize, prepare for, and ultimately conduct their operations. AI can assist with everything from planning, maintenance, budgeting, decision making, information gathering, intelligence targeting, autonomous weapon systems, drones, submersibles, you name it. To this end, it's not just about the specific weapons that AI may enable, but about the weapon systems that it can enable and ultimately give an insurmountable advantage to whichever country is at the leading edge of AI. Now, because of these dynamics, both economic and military, the technology development and associated build out of AI infrastructure globally is increasingly seen as a national security issue. It's being viewed through the lens of national security, particularly in this context of US-China competition.

Jacob Helberg, who is a former technology executive and has now been tapped for an undersecretary role at the State Department, had a quote recently that I thought was telling. He said, we've seen throughout history that countries that are the first to exploit periods of rapid technological change can often cause shift in the global balance of power. That line of thinking is what I believe is driving the thrust of US policy today on artificial intelligence.

Oscar Pulido: And when we've talked about AI in the past, I feel like we've talked about the impact on consumers. we've talked about how companies are adopting AI, but what you're pointing out is that it's governments, it's countries that are now thinking about how AI is going to influence the policy decisions that they take and how they're interacting with each other.

The last time we had you on, you talked about tariffs and trade policy and how the Trump administration is enacting some of these and the things that they're doing as they've been coming into office. What has been the Trump administration’s approach to technology?

Catherine Kress: Sure, and on your first point, I think what's interesting is that with artificial intelligence, unlike previous technological breakthroughs including GPS, and other technologies, artificial intelligence has really been driven by the private sector. It's not a government led initiative. So right now, what we're seeing is governments around the world trying to race and keep up to figure out what is the best way to both incentivize the development of artificial intelligence, as well as regulate its use, control its dispersion around the world. And so, it's an interesting balance that we're seeing between the public and private sectors. But on your question, specifically, as I said, the economic and national security concerns around AI are front and center in the US policy approach, and we see this across a range of areas.

So, under former President Biden, the US pursued this comprehensive effort, which included export controls, targeted investment restrictions, government funding and other measures to protect and extend the US lead and advance technologies. And the key competitive steps the administration took were to deny China access to, as well as the ability to indigenously make, the highest end semiconductors, which the US saw as having military applications. What's interesting is these security concerns were front and center for former President Biden through to the last days of the administration.

And in fact, one of the Biden administration's final acts on security was releasing what's called the Interim Final Rule on artificial intelligence diffusion. The rule was a bold set of steps, that was intended to, one, maintain US leadership on AI, and two, ensure that US (not Chinese) technology powers global AI development. In the words of Jake Sullivan and others, the goal was to make sure that the world's AI runs on American rails.

Now the diffusion rule sought to achieve these objectives by defining each country's access to advanced AI chips, keeping as much AI infrastructure in the US as possible, providing clarity to US allies and partners, how they could partner with the US on the global AI build out. And it sought to execute on this policy largely through companies, mostly US companies that the US designates as validated end users. Now beyond these overarching goals, the rule intended to do things like prevent the theft and diversion of AI chips, limit chip diversion, ensure that countries of concern were limited in their ability to develop frontier models and capabilities, maintain restrictions on arm embargoed countries like China and Russia, all while enabling the controlled diffusion of AI technology around the world.

Now I focus on this rule, it's received significant pushback. It's been the subject of some controversy, but implementation now falls to Trump. It has a 120-day comment period, Trump is going to have to decide whether he keeps these rules, whether he scraps them entirely, or whether he adjusts them in some way.

Our expectation is that he will almost certainly look to make his own mark, and some reports suggest that he may even look to make the rules more restrictive. President Trump, like former President Biden, sees AI as economically and strategically important. And he says the US must be at the forefront to take the lead over China going forward. So bottom line, there's going to be a continued sharp focus on AI leadership in the US policy approach going forward.

Oscar Pulido: And so, Jeff, maybe to bring you into the discussion, it sounds like there are some policy decisions that we have to wait on over the course of the next couple of months. in the meantime, as an investor, what does this mean for the tech sector? I'm hearing maybe that geographically the US technology sector could potentially benefit from this, from some of the head start that it's had, but maybe it's more nuanced than that. How do you think about the opportunities in tech.

Jeff Shen: I think it's certainly going to be very nuanced and that to your point, also highly uncertain. I like to think about it as any asset price is really going to be driven by two things. Cash flow surprises and also discount rate surprises. And I think on the cashflow surprise front, it's actually pretty important to think about how artificial intelligence is going to impact the profit, also the revenue of a particular company.

A lot of things evolving and changing to some of the things that Catherine talked about. So, I think, a particular company, as a particular stock we invest in, it's pretty important to think about how AI would actually impact, how profitable and also how much revenue surprise it would be.

And I think a lot of things are going to be evolving and changing. Now, interesting is also when we look at the tech sector specifically, demand and supply picture for the tech, stocks, they've actually been very much global for a long, long period of time. You extract revenue around the globe. get your supply chain around the globe, and that's very quickly changing as we speak, given some of geopolitical risks that Catherine very well laid out. So, what we're finding is really that technology independence, becomes increasingly more of a theme.

You actually certainly want to broaden your demand side as much, around the globe as possible. But at the same time, from a supply side, I think there's this national security concern that drives a lot of independence, and also making sure that things are nearshoring or, friendly, friendly shoring.

So, from that perspective, the discount rate, if you will, that how we are actually discount that future cash flow becomes much more complex. It used to be that this is actually not too much of a concern as long as the companies are generating this cashflow surprise. But now how we discount that cash flow becomes much more nuanced.

And know, from an ecosystem perspective, I think you certainly see a China-led ecosystem, a US-led ecosystem. And it's also important to think about that a lot of countries there are still what I would call ‘swing states’. They are trying to figure out how to align and where to align. And from an investment lens perspective, it’s also pretty important to think about that. There are a lot of nuances and a lot of uncertainty. it's not really black and white, especially for a lot of these swing state countries. So, I think, the world in technology investment is becoming much more complex and I think it's pretty important to stay current but also stay dynamic.

Oscar Pulido: Right. And just hearing you talk about, there's US Tech and China Tech, it fits this theme of geopolitical fragmentation. And you also mentioned supply chains coming closer to home. This has been another theme that we've been talking about over the last couple of years.

And Jeff, you mentioned China. I have to ask you about the Chinese AI startup DeepSeek, which made news a few weeks ago. I wonder if you can just give us an overview of what happened, how did the markets react, and how has this changed the AI race between the US and China?

Jeff Shen: So, DeepSeek released a new model that, seemed to have a pretty superior performance. But the more important thing is they claim that it only costs about $5 million to train their model. And the ongoing assumption that this type of large language model would cause around a hundred million to, close to a billion dollars. So, there's a huge cost, saving that seemed to be possible. Alongside with that, I think there's also a bunch of assumptions that we have, in a sense, that these large language model would also consume tremendous amount of energy.

So, there's a downstream supply chain implication. There's also this assumption that the US is certainly leading the AI race and here is a small startup that seemed to be able to do, a lot of things that the big US company haven't actually been able to do.

So, it was a reasonably dramatic news. US equity market actually saw about $1 trillion market cap being wiped out in a single day. And you have also seen Nvidia, went down about 18% on the single day. So, it was a dramatic event. Now, fast forward today, a lot of these losses have actually been recovered. The reason is I think when people dig a little bit deeper, I think a few of these assumptions that people have, people are starting to realize that things may not be as black and white, as people expected before.

Clearly the training costs is still very much a question mark, but also to be able to train a model like that, it still builds upon the billions and billions of dollars that's already been invested in large language models. So, it's not like you can go from a blank sheet of paper, spend $5 million and come up with a new model. So, that's one and I think it's while there are many clever things that DeepSeek has come up with, there is this expert system that they've come up with which is rather than have a generalized AI system - they call on specific experts one point at a time. So that way is actually a pretty efficient way to train the model and also to add to provide reasoning.

And so, there's some interesting and innovative approach embedded in there, at the same time, it leans also quite a bit on US technology. So, this idea that you can actually have a native developed Chinese technology completely independent of the US technology. That assumption's also not necessarily right.

If I zoom it out a little bit, it actually does highlight that from the methodology perspective, it is quite possible that you actually would see methodological diversification and there could be many different ways to develop large language model and artificial intelligence method on a forward-looking basis. So, I think that's probably here to stay. The reliance on hardware where some of the chip companies, especially in the US, certainly have a very strong positioning that's probably more likely to last on a forward-looking basis.

And the last, but not the least is, I do think that the magic will still be in the applications. How do you adopt AI? How do you use this AI? How do you use it for your advantage? To Catherine's point, it does become a bit of a zero-sum game in the sense that the people who use AI effectively versus people who don't, there will be a huge competitive advantage. So, I think long story short, is really to say that this is certainly ongoing, and I think the race is still very much open.

Oscar Pulido: And you've pointed out in the short-term, it seems like the market is back to where it was, but presumably Catherine, there are geopolitical and policy implications that are more intermediate and longer-term that come as a result of the DeepSeek news. Is that how you're thinking about it?

Catherine Kress: Absolutely. DeepSeek developments raised important economic and geopolitical questions. Economic questions that Jeff touched on around how capital intensive is the AI model really. and then geopolitical questions specifically around the US policy approach, especially when it comes to export controls. Now it's worth noting that there had been some concerns building for some time around the effectiveness of the US approach with people highlighting a number of issues like outright evasion, tech evolution and workarounds, indigenous innovation.

This is one of the reasons why we saw so many steps from the Biden administration on semiconductor export controls. Because each next step was an effort to tighten the controls, identify loopholes, fill gaps in the system.

Now I will say though, DeepSeek sparked significant questions around these issues. The overwhelming responses Jeff noted from hyperscalers, from investors, and importantly from policy makers, has really been to double down. And I think what we're going to find is that the post DeepSeek world looks a lot like the pre-DeepSeek world, but now with a much clearer competitor to the US AI ecosystem.

So, from a policy perspective, Trump, as I noted, is sharply focused on maintaining US leadership in AI. He called the DeepSeek developments a wakeup call, he said that the US needs to be laser focused on competing to win. And this is an area of continuity between Biden administration and the Trump administration. So going forward, I expect a raft of policy measures intended to do a number of things.

One, incentivize and enable the AI infrastructure build out in the US, on US soil, including by easing permitting for AI data centers, expanding energy production, loosening restrictions on AI safety and security. President Trump is surrounded by many members of the accelerationist wing of the AI movement, who generally believe that any form of restriction or regulation is stifling to AI innovation and development.

But then two, there will also be an effort to regulate AI development infrastructure abroad by imposing export controls, tariffs, and other measures. Just the other day, we saw President Trump announce that he expects potentially around April 2nd, to announce 25% or higher tariffs on semiconductor imports to the United States. And the goal here is really to reshore semiconductor production in the United States. Vice President JD Vance gave a speech in Paris the other week that I thought was telling. And in it he said, the United States of America is the leader in AI, and our administration intends to keep it that way. He stated, unequivocally that the US intends to lead on chip design, algorithms, applications, infrastructure, will not be contained by external regulation. He said that the US would continue its efforts to prevent the theft and misuse of American AI and chip technologies, working with allies and partners to close pathways and loopholes, and then he affirmed the administration's intent to concentrate the AI build-out on US soil. Going forward on this note, our expectation is that AI data centers and industrial clusters more generally are going to be considered, as I said, through the lens of national security, increasingly treated as critical infrastructure in the United States with all manner of security protections that entails.

Now a final point - and just to build on something Jeff said earlier - from a geopolitical perspective. There is increasingly the view that DeepSeek as an open-source model could widen the gap between the US and its partners and allies and the rest of the world, especially the emerging and developing world. And I say that because now there is a cheaper, high-quality Chinese model on the market that happens to come with less restrictions and regulations around its use.

So, if you're an emerging and developing market who's looking to embed AI in its economy and processes, this presents an attractive alternative to the US tech model that once again means that the global AI race is going to get a lot more competitive.

Oscar Pulido: So, I would think that investing in AI or tech is hard enough in terms of identifying a company Jeff, to talk about, what you mentioned in terms of the cash flows, identifying those cash flows and trying to value a stock. But, Catherine, you've added a whole bunch of other variables that now need major consideration when it comes to, policy decisions. You've used the term zero-sum game to talk about the race that is taking place for AI dominance.

So, Jeff, maybe to come back to you, you're managing a team that is investing and having to take all of these things into consideration. So, what are your thoughts on the biggest risks or opportunities for 2025?

Jeff Shen: Listening to Catherine, I think, certainly, our job is getting, harder. that's for sure. But it's also both the risk and also opportunities. the first point that I want to make is the sense of the country as a unit of analysis it's becoming more and more important. I think, historically for a while given the globalization trend this idea of thinking about country as a strong unit of analysis was getting weaker and weaker. But I think we're reverting back that we really need to think about, a particular government's, The national security concerns that, Catherine very nicely laid out, and also just to really think about foreign currency, for example, as a transmission mechanism to either amplify or dampening some of these policies. So, I think country as a unit of analysis is becoming more and more important.

The second one here is also. maybe we've actually been living in a world where the trending insights or the major trends have actually been continuing for a while, and I think we're entering a period we are likely to see more reversal, more contrarian movements, because there's a lot of uncertainty, DeepSeek to the point that you just raised, Oscar, that it could be a big news for a day and then things start to revert. So, I think we're going to actually, this idea that we are living in a world where, we have to watch out for some of this trending, behavior, but really to think about potential reversion as a more of a prevalent behavior in the market.

And last but not least, I do think that data and insights are becoming, more and more important in this new dynamic and complex world. I give you one specific example, what we are doing now is actually, essentially try to look at some of a policy movement that's actually coming from Washington, DC and also around the world to use policy as a pulse and then thinking about country as a unit of analysis. Then you really need to map out, both from a revenue profile perspective and also from a cost structure perspective, how global a company's profile is, and also where the geographic locations are from a revenue and from a cost perspective.

Then you essentially use the policy as a pulse to see what other shocks can be. So, to the extent that you can go for a deeper analysis to adapt to some of these policy dynamics, there's actually a lot of opportunities from an active management perspective. So, I think, the job is getting harder, I think the possibility to generate alpha in this new environment, could certainly also be richer.

Oscar Pulido: Right. It's getting harder, and you mentioned country risk, being a unit of analysis that it is reintroducing, so I hope you have a geopolitical expert on your team, Jeff, and if not, maybe Catherine, you'll make yourself available to talk to Jeff. it's great to have you both back on.

We talk a lot about the five mega forces, are at play in the global economy. We don't often talk about how they intersect, but it's been great to hear about how geopolitics and AI are really, at that intersection and what the risks and opportunities are. Thank you both for sharing these insights and thank you for doing that on The Bid.

Catherine Kress: Thanks for having us. 

Jeff Shen: Thanks Oscar!

<<THEME MUSIC>>

Spoken disclosures at end of each episode:

This content is for informational purposes only and is not an offer or a solicitation. Reliance upon information in this material is at the sole discretion of the listener.

For full disclosures go to Blackrock.com/corporate/compliance/bid-disclosures

MKTGSH0225U/M-4263274