whizzy1919@gmail.com
We are maximally dependent on the AI superiority of the USA. But we can and must do something about it, demands change expert Kishor Sridhar
Europe can drill but it doesn’t yet control the pipelines
Almost everyone now uses AI solutions that are based in the USA. There are promising solutions in Europe. We don’t lack talent.
The best known are probably the German one Aleph Alphawhich develops large language models, or the French provider Mistral AI, as well as the world leader in AI translations DeepL from Cologne.
Or Black Forest Labs, which is successfully working on generative image AI. Also exciting is the French start-up LightOn, which develops AI solutions for knowledge work that are particularly suitable for internal company data. However, growth in companies primarily comes from meaningful integration into production processes.
That’s exactly what Edge AI does. Edge AI is AI that runs directly in machines or devices. Like a small refinery at the production site, data remains in the system instead of flowing through international pipelines.
This is an effective and safe application of AI, and companies such as the machine tool manufacturer Trumpf, Deutsche Bahn with intelligent condition monitoring of trains and infrastructure, and the automation specialist Festo are successfully demonstrating it.
Kishor Sridhar is a respected consultant, keynote speaker and author specializing in change management, leadership and digitalization. He supports managers in transformation processes and teaches at the ISM in Munich. He is part of us EXPERTS Circle. The content represents his personal opinion based on his individual expertise.
Germany and the EU must take concrete action now
So in Germany and Europe we can do AI. But if we want to become more independent in Europe, we must not ignore developments. We need economic and industrial policy measures.
1. Establish an EU-first procurement logic
The EU must create demand. When tendering, public clients should place greater emphasis on criteria such as data sovereignty, European legal binding and Security architecture demand. This creates a demand stimulus for providers who meet European standards.
2. Massively expand computing power
We need high-performance computers with bundled computing capacity. Community-funded programs such as EuroHPC and planned European AI Factories are a first step and should be further expanded.
3. Secure growth capital
Through targeted growth funds, we can prevent promising AI companies from migrating to the USA at an early stage instead of developing in Europe.
4. Less regulation and still protection
The EU protects us from uncontrolled growth. But that slows down innovation. That’s why we need regulatory sandboxes, i.e. legally defined experimental spaces in which companies new technologies can be tested under official supervision for a limited period of time before all regular regulations take effect.
Each of us can do that
But each of us can also contribute privately to freeing ourselves from dependence on the USA. You can apply the same five-step plan that we use in German companies to become more independent when it comes to AI:
1. Gain transparency
Make a checklist: Which AI services do you regularly use? Where is your data processed? Anyone who knows their digital infrastructure can make more conscious decisions.
2. Actively test European alternatives
Use DeepL consciously for translations. Try chat interfaces, text assistants or business solutions that run on Mistral or Aleph Alpha.
3. Run a local model
With tools like Ollama or LM Studio, AI models can be operated directly on your own computer. This means you can run a small AI refinery in-house and sensitive data won’t leave your system.
4. Check your cloud usage
Don’t rely exclusively on US cloud services such as Google Drive, iCloud or OneDrive, but consciously move some of your data to a European cloud provider.
5. Diversify search and browser structure
The usual search engines and browsers have now become AI data octopuses. Google and Co. aren’t everything. Consciously use European providers with better data protection. Every single step contributes to more independence.
The AI race is not lost (yet).
In Europe and especially in Germany, we have sufficient industrial depth, strong medium-sized companies and enormous research expertise in artificial intelligence. In contrast to oil, we have the potential right here.
However, we will lose the AI race in Europe if we are too comfortable or too slow. It is a question of decision and determination at all levels whether we want to become fatally dependent on the USA or use our potential for a free future.
