Some experts believe that corporate leaders who are more worried about tariffs than the future of AI are worried about the wrong thing.
“While everyone is focused on tariffs, I think that the next frontier of geopolitics is AI,” says Nikolaus Lang, global leader of the BCG Henderson Institute. “This may sound bombastic, but AI is creating a new geopolitical order that we are living in.”
As it stands now, the US and China have defined their dominance in AI ownership, development, and investment.
According to research from BCG Henderson Institute, of the total capitalization of the 1,000 largest public tech companies in the world, the US accounts for $24.7 trillion (with a “t”)—five times the share of the Asia-Pacific market and 18 times that of the European market. Between 2019 and 2024, gen AI startups had received $65 billion in private investment, according to data from BCG. The US also has the world’s largest data center capacity and has secured the most reliable access to chips.
Meanwhile, China is working to close the gap, with the second-largest pool of AI researchers (around 18 percent of the world’s total, compared to 60 percent in the US), the production of 15 percent of the world’s notable models, and around $110 billion in government venture capital investment in AI since 2019, according to research published in February. Through strategic development and investment, Chinese models have now effectively reached parity with frontier alternatives, according to experts at the BCG Henderson Institute.
At this point, it’s undeniable: AI technology is the race, entire global markets are the prize, and the US and China have established staggering gaps with the rest of the world when it comes to development. But development isn’t the only way to stake a claim. While research indicates that most countries cannot become self-sufficient suppliers of AI, business and government leaders can shape future global markets by applying AI to their markets. That leaves players around the world jockeying for position to become leaders in adoption.
“When it comes to owning AI technology, most countries are out of the race already,” says Sylvain Duranton, global leader of BCG X, the tech and design unit of Boston Consulting Group. “That’s a very different type of challenge than the consumption one, which to me is the most critical. Consumption will, in the end, be the name of the game because companies that jump and embark on the AI journey sooner will be better, stronger, and more successful at adaptation, strengthening their competitive position. They will be the future world leaders of their sectors.”
There are opportunities for private companies to apply AI and generative AI to call center processes, sales processes, coding, the design of factory processes, and the optimization of R&D processes. But—this is a big “but”—it requires that company leadership and IT decision-makers invest the time and money to understand the tech and its application for a successful adoption, according to Lang.
Even countries or companies that make financially modest investments in AI adoption can have disproportionately large impacts.
“A new landscape is emerging where smaller countries are gaining influence through targeted investment, strong partnerships, and focused policy,” says Vaishali Rastogi, global leader of BCG’s Technology, Media and Telecommunications practice.
New Regulatory Realities
Since 2023, legislative mentions of AI rose by 21.3 percent across 75 countries. Adoption largely depends on regulatory environments, as policy can bolster or dampen enterprise use of new tech. Policy also dictates the barriers between operations, production, and workflows across borders.
“There’s a risk when competition becomes siloed,” says Rastogi. “Without coordination, we may see a fractured AI landscape with incompatible standards, duplicated efforts, and uneven access.”
According to reports from the BCG Henderson Institute, around 40 percent of large companies have at least some of their tech teams in more than 10 countries. As regulation occurs in silos around the world, business decision-makers are suddenly faced with those challenges, including the unavailability (or banning) of certain technologies produced in specific countries. For example, Italy has banned all Chinese AI platform products. Russia, China, and Iran are restricted from receiving advanced chips from the US. These kinds of regulatory and practical subtleties between countries mean that a company could reasonably be faced with building different, isolated tech stacks in each country where they have a presence.
“Corporate leaders are in a very challenging situation,” Lang says.
“Today, when you do a vehicle navigation system, you have to be sure that if you use it in China, it uses digital maps and the data is being saved on Chinese cloud computers. If you use that navigation system in Europe, you need to make sure that the data is compliant with EU regulatory law. And if you want to use that navigation system in the US, you have to comply with the new ICTS regulation, starting on January 1, 2027—meaning that if you happen to have Chinese code in your navigation or your driver systems, after 2027, you can’t sell the car. So the reality is that if I’m a car manufacturer, I already have three different regions where I need to have three different tech stacks to satisfy regulations,” Lang continues. “And that’s just the beginning. You need to understand the realities. You need to be aware that the one-size-fits-all approach and the single global product will not work, and that you will need to adapt.”
Because of the new geopolitical realities of technology and AI, business is moving from an old paradigm that allowed companies to concentrate resources, production, teams, and supply chains where headquarters are managed to a new normal—where three or four of those global locations exist within different regulatory bodies.
Building Something That Lasts
Lang remembers a time not too long ago—only about a year and a half ago—when the question about AI was: Will it last? “Now I think everyone understands something really is coming,” he says.
The reaction within the business world has been a flurry of interest and toe-dipping into AI strategy, but also a lot of trepidation. The sentiment of “we want to do something, we just don’t know what,” reverberates through many a conference room.
Duranton encourages that “something” to be just that—something. An effort into AI integration, even if it’s not perfect.
“You have to decide to move,” he says. “And why am I saying that? In the last six months, I’ve seen some business people saying, ‘You know, we were all talking about generative AI last year. Now everyone is on autonomous agents in one year, and gen AI in two years … should we pause and wait for the technology to be ready and steady before we do things?’ My advice is no. You have to adopt now because there won’t be a pause in the foreseeable future. So you have to do things now and be behind by as little as possible.
In his experience, the initial shift within an organization is the super heavy lifting.
“Even if you do it with a technology that you know will be obsolete in 18 months or two years, that heavy lifting is done. And provided you’ve deployed your solutions in a way that is quite modular, upgrading will be the easy part,” he adds.
Integrating AI is just the first step. Using it successfully requires business and IT leaders to stay nimble, using the tech to navigate shifting political regulations and new technological advances as they come.
“Embracing AI means accepting that it’s here to stay, and that brings both opportunity and responsibility,” says Rastogi. “Leaders must invest in regional capacity, support open and flexible infrastructure, and manage workforce upskilling to build readiness. At the same time, governance frameworks must ensure responsible use without stifling innovation. Getting this balance right is how we make AI serve economic and societal priorities, not work around them.”
Discover more about BCG here.
