The EU AI Act applies to providers placing on the market or putting into service AI systems or general-purpose AI models within the EU, and deployers of AI systems within the EU. Notably, it also applies to providers and deployers outside of the EU if the output of their AI system is within the EU. This means that regardless of where a company is based, if its AI system impacts users within the EU, it must comply with the EU AI Act.
“Output” is not explicitly defined within the EU AI Act. However, one may refer to the definition of “AI system” within the Act. This refers to outputs such as text, videos, images, predictions, recommendations, and decisions that influence the physical or virtual environment.
The EU AI Act has also introduced certain obligations for importers and distributors of AI systems, product manufacturers, and authorized representatives of providers.
Why does the EU AI Act have such a broad scope, even affecting providers and deployers outside the EU?
The EU considers AI to be a powerful tool, which therefore requires the protection of its users. The solution, as seen in the text the Act, is an Act that covers any AI system or general-purpose AI model impacting EU users. The EU Act in fact aims to:
- Protect fundamental rights: the Act ensures that AI used within the EU respects its cornerstone values. These are the values of privacy, non-discrimination and human dignity.
- Ensure fair competition: all players operating within the EU market must follow the same rules. This applies whether they are based in Berlin or in Boston.
- Lead global standards: the EU’s approach aims to champion ethical, human-first AI development, setting a blueprint for worldwide regulation.
Will the EU AI Act kill AI innovation in Europe?
There has been a lot of talk on the adoption of the EU AI Act and it killing AI innovation within Europe. However, regulation will not necessarily stifle innovation. In fact, it can be seen as guiding the industry in the right direction as:
- Clear rules allow may allow for faster progress. Having one rulebook that applies across the EU allows for less red tape and less cost in navigating multiple regulations. This creates overall simplified compliance for global companies.
- The GDPR effect. Just as the GDPR shaped global privacy standards, the AI Act could become the worldwide blueprint for ethical AI.
Conclusion – Why AI “output” matters for compliance
In summary, when thinking about the scope of the EU AI Act, one must keep in mind that the EU AI Act cares about impact, and not where your company is based. The bottom line is that if your AI “output” reaches EU users, the EU AI Act applies to your business.
Have any doubts as to whether the EU AI Act applies to your business? Contact Blue Arrow today for a consultation!
