EU red tape is killing Europe’s AI boom, but it’s fixable
Europe’s AI ambitions are right, but the current regulatory approach risks making innovation harder, not safer. As implementation begins, the key question is how to remove duplication and make the rules work in practice — without lowering standards.
Europe is known for setting the global standard in safe, high-quality products. At the heart of this success are its manufacturing industries — machinery, medical devices, automotive and radio equipment — which support 30 million jobs across 2.2 million companies and form one of Europe’s strongest economic pillars.
Countries like Finland show what this looks like in practice: world-class engineering, advanced manufacturing, and a fast-growing AI ecosystem ready to scale.
But right now, Europe risks holding that progress back.
The ambition behind the AI Act is right. Safe and trustworthy AI is essential and widely supported by industry. The problem is not the goal, it is how the rules work in practice.
Today, companies can face overlapping requirements, effectively navigating two parallel compliance processes for the same product. In reality, this often means doing the same checks twice.
For companies developing AI-enabled products, from smart factory systems to medical devices, this creates duplication, delays and uncertainty. It does not make products safer. It just makes innovation harder.
And it comes at a real cost.
Introducing AI into a product can require between €320,000 and €600,000 upfront, plus ongoing compliance costs. For many SMEs and scale-ups, that is enough to put projects on hold or even cancel them entirely.
A single, streamlined assessment pathway would not weaken the AI Act.
We are already seeing this happen. Promising innovations are being dropped, not because the technology fails, but because the system is too complex to navigate.
At a time when Finland and Europe are investing in digitalisation, productivity and the green transition, this is a risk we cannot ignore.
The frustrating part? The solution already exists – yes and it is already embedded in existing frameworks.
In sectors like aviation and automotive, AI requirements are built into existing product safety frameworks. One process. Clear rules. No duplication.
It works and it keeps safety standards high. There is no reason not to apply this approach more broadly.
A single, streamlined assessment pathway would not weaken the AI Act. All core requirements on transparency, human oversight and data quality would remain fully in place. It would simply make them easier to apply.
And this is not just an industry concern, it is already recognised at national level. For example, the Finnish government already recognises that the Commission’s simplification proposals do not go far enough. The next step is to turn that position into active influence at the negotiating table — and deliver simplification that genuinely reduces burden without compromising safety.
Finland has a strong voice in shaping Europe’s digital future. Turning that position into concrete influence now will be critical not only for Finnish industry, but for Europe’s competitiveness as a whole.
The global context is moving fast. Europe already lags behind the US and China in AI investment. Every additional layer of complexity makes it harder for companies to scale, invest and compete. At a time when Europe needs growth, jobs and innovation, this is the last thing we should be doing.
We still have a window to get this right. And we must do it.
As the AI Act moves into implementation, the focus must be on making it work in practice. Not by lowering standards but by removing unnecessary duplication that slows innovation.
Europe and Finland have everything needed to lead in applied AI.
What is needed now is a regulatory approach that keeps pace with innovation and investment, enabling companies to scale and compete globally.
By Cecilia Bonefeld-Dahl, Director-General of DIGITALEUROPE, and Taina Susiluoto, CEO of Technology Industries of Finland
Further information:
