America can switch off the world’s AI. Europe must adapt

Artificial intelligence started as an invention, continued as a business model, and ended up this spring as a weapon of mass destruction.

Since Anthropic – a leading developer of AI platforms and competitor of ChatGPT or Gemini-producers introduced its most powerful model Mythos – the world has become unsafer.

Mythos is capable of hacking sensitive software worldwide and AI has now officially become both a national security threat and a defence asset. The problem is: Europe’s capabilities are nowhere close.

‘Europe is becoming a digital colony between two AI empires’

The introduction of the model was just the beginning: Project Glasswing – the initiative that was founded to research and contain the risks associated with Mythos – does not include any European actors.

And this month’s decision by the US administration to withhold access to Mythos for non-US nationals was the first of its kind – the proverbial AI kill switch in action.

It showed – if the White House wanted to, an estimated 88% of the world’s population could lose access to American frontier models overnight. China is projected to obtain similar capabilities within a year. Europe is becoming a digital colony between two AI empires.

For Brussels it’s time to act. Though we have fallen far behind, a combination between sovereignty, scale, and reforms is our only chance to catch up. Until then, we need a smart strategy of cooperation that will keep our economy alive.

Unlocking Europe’s scale

First and foremost, it is about scale. This is not the race we have chosen, but this is the race we are in.

In this race, compute power and innovation are driven by enormous investments. Trillions are more realistic than billions. Equity of this magnitude will not come from public pockets. It will need to be private.

The Commission must convene an emergency summit with the leading European businesses. We need pledges that are substantial and reliable. The big bosses must realize that it is about their own survival, and it is about the future of Europe as a business space that can defend itself.

At the same time, the EU capital market Union and the single digital market must be completed in a fast-track mode.

Scale is not only money, it is also the size. Our economies of scale are economies of Europe. National solutions will not do the trick. We need French-German Gigafactories and European space constellations. We must use the asset that we have – our common market and our common synergies.

Multilateralism against digital imperialism

Second, the Commission must marshal international partners. Pool AI compute – capabilities to process data for AI – beyond European borders towards a consortium of middle powers.

The EU is the only superpower capable and willing to use multilateralism to confront digital imperialism. The EU’s international digital strategy already sets out the framework for the Commission to act on digital foreign policy. It proclaims “global partnerships” in digital cooperation as its goal. Now it is time for Europe to lead on these efforts.

Finally, and in parallel to this, the Commission should build a new relationship with the US. It is a hard pill to swallow: whether we like it or not, in the medium term, Europe will depend on American compute and chip infrastructure to grow its own providers at the application layer.

We have leverage, like the Chip printing technology – the only in its kind Dutch company ASML. But this leverage is limited compared to the powerful US capabilities.

What Europe needs is a path of strategic parallelism: cooperating with Washington where we must, while building capacities and infrastructure that will one day let us stand alone.

It may surprise some deregulator freaks, but regulatory standards are of help. They anker American technologies in European trust and transparency. These are the only foundations which can secure us in the current world of threats and coercion and can secure American tech an access to our markets.

No AI strategy without public trust

The United States is seeing the so called techlash – a growing mistrust towards AI infrastructure among citizens.

We will only avoid similar reactions if innovations in energy efficiency or water use become an integral part of this new transatlantic cooperation.

Add to this social redistribution of resources and profits – so that citizens are not the ones squeezed by AI giants. Only if we achieve public acceptance of high speed investments those investments will be politically and economically fruitful.

Those are uncomfortable messages of truth. Rallying capital, herding allies and bargaining with an unpredictable partner all cut against our European instincts for caution and consensus.

But the American kill switch has already shown that we entered the uncharted AI territory. The only open question is whether we Europeans will manage to take our destiny in our own hands and turn in a geopolitical technology player.

The alternative is that we wait for someone else to switch us off. And this we should never accept.

Dr. Sergey Lagodinsky is a politician of the german party “Bündnis 90/Die Grünen”. Since the 2019 European elections, he has been a member of the European Parliament and was also elected Deputy Chair of the Greens/EFA Group. He has been Chair of the EuroNest Parliamentary Assembly, a member of the Committee on Foreign Affairs, the Committee on Legal Affairs and the Delegation for Relations with Belarus. He is also a deputy member of the Committee on Industry, Research and Energy, the Delegation for Relations with the United States and the EU-Turkey Joint Parliamentary Committee. He is Lawyer and author.