Alberto Alemanno: How AI Will Reshape European Politics
The HEC Paris professor on the EU’s dependence on US technology, Big Tech’s influence on policymaking, and the potential rise of anti-AI parties. An interview by Stefano Feltri
Can Europe regulate Big Tech while relying on it for essential technologies? Alberto Alemanno, Jean Monnet Professor of European Union Law at HEC Paris and founder of The Good Lobby, examines how that dependence constrains European policymaking.
This conversation with Stefano Feltri was recorded for IEP Bocconi’s Policy Insights podcast on the occasion of the institute’s Annual Event in Milan on 18 November 2026, “At the Table or on the Menu? Europe, Middle Powers and International Order”. Alemanno moderated the panel “Technology and Industrial Policy: The Race for Digital Sovereignty” and contributed the commentary “Digital Sovereignty Is About Power, Not Self-Sufficiency”.
He discusses the pressure on EU regulators, choices on AI investment and Big Tech’s influence on public debate. He also anticipates the emergence of anti-AI parties, bringing together concerns about jobs, cognitive autonomy and the environment.
After almost two years of Donald Trump in the White House, how do you assess the European Union’s independence in regulating the digital industry, especially American Big Tech? Has the EU succeeded in applying the DMA, the DSA and other rules, or have its regulators been constrained by the geopolitical imperative not to upset the US president?
It took five years for the European Union to develop a digital rulebook capable of taming Big Tech. When Trump returned to the White House, it coincided with the moment when the Digital Services Act and the Digital Markets Act entered into force. This immediately had a chilling effect on the authorities, on DG Competition and on DG CONNECT, which is responsible for digital policy.
There have been public acknowledgements that the authorities were not going all the way, as was also acknowledged in the Turnberry deal. The American demand was clearly not only to enforce less, but also to water down, for instance, the Digital Fairness Act, which is the missing piece of the puzzle. The gaps left by the Digital Services Act are supposed to be filled by the Digital Fairness Act. So there has been mounting pressure on the European institutions not to implement their rulebook fully.
The Americans have come to expect an easy pass when entering and remaining in the European market. But this is now changing very suddenly as a result of the midterm elections and the blue wave, which has clearly weakened this administration’s ability to call the shots. It may be too early to say whether this alone will give the European Union, and particularly the leadership of its institutions, enough political courage to go all the way.
It is not entirely correct to say that the EU has not applied or enforced its digital rulebook. There have been a few instances. There is a major Google investigation that could theoretically lead to a record fine. But we have seen the institutions tactically delaying it. We have heard about a lot of pressure from Big Tech itself, but also from the American Chamber of Commerce and other interest groups that clearly do not want the decision to be taken.
Will the EU seize this window of opportunity to go all the way, apply its rulebook and see what happens? There are ideas circulating that the EU should become a role model in applying antitrust decisions. These are not trade measures, as the US administration often likes to depict them; they are ways to regulate the proper functioning of the European internal market.
But this is still largely a theoretical debate. The only person speaking frankly about it is Teresa Ribera, the European Commission’s executive vice-president responsible for competition and the clean transition.
You mentioned the midterm elections. How do you expect Europe’s approach to digital regulation to change over the next couple of years? Are there specific measures that could take a different path now that Trump is a lame duck in terms of his political power and, perhaps, his international influence?
Will the EU find the political bravery to push back against this administration, now that it could take advantage of the president’s lame-duck position? Obviously, Trump is not a traditional lame duck. He will remain in control of the US administration, perhaps not of the two chambers of Congress, but he will still be able to make it difficult for the EU to discharge its own duties.
I think the high point was the political interference exercised by Elon Musk through X. We have also seen Meta withdrawing its commitment to countering disinformation. At that moment, the EU considered relying on the so-called bazooka, the Anti-Coercion Instrument, which would have been the cherry on the cake. The political conditions do not seem to be fully in place yet, but it remains a possibility.
The only jurisdiction in the world that has used a kill switch against one of the Big Tech companies is Brazil, where X was suspended for 24 hours. That alone prompted a very significant reaction from the company. So there is a precedent the EU could potentially use.
Suppose that, in the remaining two years, we see another instance of economic coercion or a military threat, as in the case of Greenland. We could imagine US Big Tech being on the receiving end of that process. But when it comes to day-to-day decision-making, I doubt whether the EU will use its full powers to limit these companies’ influence. At the end of the day, the dependence is too significant.
As we have seen in recent months, when a new European tech company enters the market, the last mile, commercialisation, still has to rely on some US technology. W Social, the supposed European Twitter, is one example.
It is not simply that the US industry is limiting our ability to apply the rules. This unprecedented overlap between Big Tech companies and the US administration is creating a situation in which public authorities are extremely afraid of testing the limits of the system.
I assume this is also true of artificial intelligence, where Europe has no major AI company, with the partial exception of Mistral. How should Europe approach the choice between trying to compete with the US along the same lines and taking a different approach, as China has tried to do? Could it focus on industrial and practical applications, or open-weight models, rather than very large language models? Where do you see the real opportunities for the EU?
The debate about how the EU should move forward on using, developing and supporting AI is alive and kicking. At one end of the spectrum are the enthusiasts, the AI accelerationists, who say we need to develop our own capabilities and our own frontier models. Their assumption is that only by doing so at the deepest level can we remain in control of our technological agency and therefore be sovereign.
At the other end are those who say we have lost that battle: it is too late and it would take too much time. This is a battle we should not pick. Instead, we should be good at using AI and increasing its uptake, while reserving areas for European preferences along the technology stack.
I believe in a middle position. We should avoid ending up without models of our own. We need to invest in them, but that is a long-term project. At the same time, we should be very good at using AI and developing applications that make its use widespread, with more geographical control and our own computing capacity. When you break down the AI stack, you can see that our weakness is not universal.
There are areas where we have capabilities. Chips are largely a US-Taiwanese business, but we have the Dutch company ASML, which gives us a competitive advantage we should preserve. When it comes to turning these capabilities into applications, I think that is where the sweet spot lies.
Our cultural and linguistic diversity might give our researchers the ability to create something that, in turn, creates dependencies: strengths Europe could have vis-à-vis the US or Asia. We need to find the right formula, with different ingredients, rather than throw the baby out with the bathwater by saying either that we are going to win the battle or that we are not even going to try.
We need to become much more sophisticated about identifying the areas of the AI stack where we could be competitive and reinvest in those, while investing much less in areas where dependence is effectively built in and we cannot overcome it.
I am very hopeful that this conversation will not remain a purely European one. The challenges Europe faces in the AI stack and in maintaining technological sovereignty are the same as those faced by Canada, Japan or South Korea. That is where Carney’s approach, coming together and building alliances that might also include supply-chain protections, could provide an opportunity.
In the meantime, as Europe moves forward, we see America developing Pax Silica. These are projects aimed at countering China and building resilience along AI supply chains, which already have the support of individual member states. It is not the Union signing up, but some of its members, together with other former US allies and the US itself.
There is an ambiguity here, which I think reflects each government’s ambivalence about whether to decouple from the US in technology or try to keep the best of both worlds. That is the situation we are in: looking for new partners while remaining attached to where we come from.
These companies also have huge amounts of money to spend on influencing regulation and shaping the political process, sometimes overriding voters’ preferences. As an advocate of good lobbying and transparent decision-making, how do you think we can ensure that these strategic choices are made transparently, without being captured by private interests?
Big Tech is the worst in class when it comes to lobbying methods. These companies do not simply rank very highly in the number of meetings they hold or the policy papers they produce to influence decision-makers. They engage in what I call upstream lobbying: they change the informational landscape in which ideas are generated.
We see academic chairs being funded and more and more efforts to change the landscape in which decision-makers position themselves. This translates into public administrations becoming entirely dependent on their technology. When you look at these structural factors, you see that Big Tech is doing more than capturing our decision-making processes. It goes as far as challenging our individual human agency and our cognitive processes.
This is a challenge on a unique scale. It cannot be curtailed by an individual European company or organisation; it needs a solution through the political process. I think many of the protests we have seen across Europe and elsewhere convey the idea that AI is not necessarily what most people want today. Political forces able to channel those voices will probably gain electoral support in the future.
We should not rule out the possibility of single-issue parties, as we have seen in the past around animal protection or the climate.
An AI party, or an anti-AI party?
Anti-AI parties, I think, are the next frontier in politics. You could mobilise millions of voters who, for a variety of reasons, are concerned about the labour market or the possibility of being cognitively captured. Others are concerned about the environment and data centres, which are just the tip of the iceberg.
These groups might coalesce into anti-AI parties, which I think could play an important role in moderating the enthusiasm we have seen across the political spectrum. Politicians say we are going to miss the train, that we need to invest and position ourselves, without really considering the trade-offs within society.
The big debate is this: if all these investments go into the technology stack, what will that cost welfare and social spending, at a time when we are also doubling down on investment in defence?
All of this is coming together at the same time. Politically, it could play out in very different ways. But America and the American political debate are clearly ahead of us, because AI penetration and the pushback it has generated are much more significant there than in Europe.
We already have plenty of indications that what is happening in America will happen in Europe. It is just a matter of time. It takes longer to coordinate all these movements.
IEP Bocconi does not express opinions of its own. The opinions expressed in this publication are those of the authors. Any errors or omissions are the responsibility of the authors.