AI is boosting the US economy: I've never seen anything like this before

Rather than commenting on a series of recently published macro figures, I want to share two important messages today. The first, which I also wrote about last week, is that an unprecedented AItree is unfolding in the US. Never before have I seen developments in a single sector impact macro figures so strongly in such a short time. The rest of the world is also benefiting. Europe seems to be ahead of the curve in terms of AItree to lag far behind.

My second message is that inflation will fall structurally below the central banks' target over the course of this year. Disinflationary forces are much stronger than inflationary ones. Therefore, interest rates are more likely to fall than rise.

Unprecedented AI-tree in the U.S
Despite Donald Trump's import tariffs, the American deficit in international goods trade has increased in 2025. Perhaps this is a sign that these tariffs have less impact on trade than you might think. Incidentally, these tariffs are certainly not entirely ineffective. The enormous deficit the US is recording in bilateral trade with China has decreased by about a third in 2025 compared to 2024. Compared to 2018, the US deficit with China has more than halved, although it is likely that many Chinese goods will still reach the US market, albeit via a detour. I'll leave that aside for now. The fact that US car imports have fallen sharply in 2025 also demonstrates the effectiveness of tariffs.

So how could the US goods trade deficit have increased? In my opinion, it has everything to do with the unprecedented AItree In the US. Imports of all things AI-related have grown dramatically. What's disturbing for us Europeans is that we don't see this at all, or much less so, in Europe. This is most clearly reflected in Taiwanese exports. These are heavily concentrated in semiconductors, which are needed for AI. The first graph speaks volumes. Admittedly, Taiwanese exports to the EU are growing rapidly, but are much smaller than those to the US. The US bilateral trade deficit with Taiwan has increased from $74 billion in 2024 to $147 billion in 2025. Between 2002 and 2018, this deficit fluctuated between $10 billion and $15 billion annually. These are values, so including price increases, but in terms of volumes, the picture is undoubtedly also spectacular.

The following two bar charts show trade between the US and China, respectively Taiwan. I took them from Bloomberg.

Would those guys…?
What does all this mean? A few things. First, Europe is lagging far behind. I hear bureaucrats in Brussels, and by the way, those in The Hague too, constantly arguing that we in Europe need to become less dependent on American tech companies. Would these people look at these kinds of figures? Would they realize how far behind Europe is? Would they ever do some critical self-reflection and ask why? I'm afraid not.

In the short term, that AI-tree the American economy a strong boost that also boosts activity in the rest of the world, for example in Europe. In that AI-tree We must distinguish several phases. So far, it has mainly focused on building AI capacity. Two scenarios are conceivable: a positive and a negative scenario. In the positive scenario, the AItree Moving smoothly to the next phase, where AI applications will significantly boost labor productivity. The fact that Europe is building much less AI capacity doesn't mean we won't find applications. Higher productivity growth is certainly possible here, but we will remain dependent on the US.

However, an alternative scenario is conceivable. In that scenario, a significant overcapacity of AI is currently being built up in the US. Overcapacity always ultimately leads to sharply declining prices. The companies currently investing the most will then be hit hardest. This will lead to significant price corrections on stock markets. In this second scenario, the outcome is ultimately positive, because AI applications will also lead to a significant improvement in productivity growth.

The improvement in the US economy is clearly reflected in the development of manufacturing output. Production growth has now risen above 2%. The last time the growth rate of manufacturing output in the US exceeded 2% was several years ago. Trump will undoubtedly claim this as a success of his policies.

Statistics Netherlands (CBS) published figures today on investments in tangible fixed assets in the Netherlands. Not exactly something to cheer about. In December, the volume was 2,6% lower than a year earlier, even though December 2025 had one more working day than December 2024. Statistics Netherlands writes: "Investment was mainly lower in other road transport (trucks, trailers, vans, etc.) and aircraft. However, investments in passenger cars, buildings, machinery (including defense equipment), and infrastructure were higher than in December 2024." And this is duly noted.

Structurally lower inflation
We already knew that inflation in the eurozone fell to 1,7% in January and to 2,4% in the Netherlands. Based on the European inflation measure, HICP, our inflation was only 2,2% in January. In the US, inflation fell from 2,7% in December to 2,4% in January. This is clearly higher than in the eurozone and also above the Fed's target. But it is still much less than feared and predicted by many economists when Trump announced his import tariffs. American quants try to filter out the effects of these tariffs from the figures and then conclude that inflation is meeting the Fed's target. These effects of the tariffs are temporary. In the following graph, I've drawn a vertical line in April of last year when Trump announced his tariffs. The inflation-raising effect has been much less than expected so far. A few American economists, who are close friends, predict that these inflation-raising effects are yet to come. Perhaps they are right. But I suspect they're making a classic mistake that I've often made too. If your prediction doesn't pan out, you argue that the numbers are wrong or that you'll be proven right in the near future. That's usually a mistake...

As I wrote above, I think inflation will continue to decline, both in our country and in the US. I have a few arguments for this. First, global food prices have been falling for months, albeit gradually, and this is putting downward pressure on food price inflation, albeit with a lag. Second, global oil and gas prices are now lower than a year ago. A US attack on Iran could throw a wrench in the works, but there is currently more than enough oil and gas in the world, so prices are more likely to fall than rise over the course of the year. Third, the world is currently characterized by industrial overcapacity, which means prices for industrial goods are also more likely to fall than rise. Fourth, wage growth is moderating in many countries, causing inflation in services to fall. And, to top it all off, AI applications will stimulate productivity growth, which will lower inflation.

If I'm right in my prediction that inflation will fall, then interest rates will also fall. The Fed will certainly lower the official interest rate further, and the ECB will do the same if it becomes clear that European inflation is structurally falling below its target. Capital market interest rates will also fall when it becomes clear that disinflationary forces are winning.

Closing
I repeat my two messages from today. First, the figures regarding international trade confirm that an unprecedented AI boom is currently underway.tree is happening in the US. Europe is lagging far behind.

Second, the January inflation figures suggest that there is strong downward pressure on inflation. My prediction is that inflation will continue to decline and will fall structurally below the central banks' target.

Hans de Jong

Han de Jong is a former chief economist at ABN Amro and now a resident economist at BNR Nieuwsradio, among others. His comments can also be found on Crystalcleareconomics.nl

© DCA Market Intelligence. This market information is subject to copyright. It is not permitted to reproduce, distribute, disseminate or make the content available to third parties for compensation, in any form, without the express written permission of DCA Market Intelligence.