The Roaring Twenties of technology regulation are dawning: For the European Union is no longer reacting to developments with its regulation, but is proactively presenting guidelines. In doing so, it could set global standards.
Setting global standards is typically a complex matter. As a result, harmonisation processes require negotiation and perseverance. This is particularly unfortunate for dynamically developing areas of regulation and leads to legislators looking for shortcuts.
The European Union’s approach to digital policy is exemplary in this respect. For years, the EU has increasingly pursued a unilateral strategy. It creates rules to form a global standard and encourage imitation. Put simply: What is suitable for the European internal market must eventually be considered a solution worldwide. This pattern was particularly evident in the case of the General Data Protection Regulation.
At the end of 2020, the EU repeated this approach with its proposals for the Digital Services Act, Digital Markets Act and Data Governance Act. At first, this sounds like a free competition of regulatory ideas. This is also the case at the outset because the EU cannot force non-European states to adopt their standards but can only propose such standards – with all the available soft power. The economic relevance of the EU internal market should not be underestimated. It is crucial for companies worldwide and acts as a catalyst for European standards. Companies align their products with the often high standards – and thus also shape practices in unregulated or less regulated third countries.
The EU bases this strategy of “upward convergence” on two premises. First, the EU assumes that European legislative acts can be transplanted to other countries and continents. This in itself is debatable. Above all, however, the EU wants to take decisions on the content of its legislation to the rest of the world because the various legislative acts express the European understanding of fundamental rights. Such export is not a priori illegitimate, but it depends on the need for the concrete “product”.
AI: A Regulation “into the Open
This is especially true for the regulation of future technologies. Here, the guard rails have often not yet been set or are at least in flux. The path dependencies are small and leave room for receptions. Against this backdrop, the EU is again using its first-mover advantage. In April, the EU Commission presented its proposal for regulating artificial intelligence particularly effectively. The proposal is a regulation “into the open” in several respects.
First, the proposal offers the first coherent approach to an AI regulatory framework. Artificial intelligence applications are divided into four risk categories – and regulated according to these categories. Such a risk-based approach is practical – some would have liked to see a similar policy for data protection. Moreover, the proposal is also characterised by an open approach to substance.
Regulatory Risk Moving Target
The EU Commission is only too aware that the speed of development of modern technologies is rapid – and that regulating a moving target is, therefore, complex and presupposing. There is a possible risk of producing legal uncertainty and structurally impeding innovation. However, waiting and watching for a long time is often not the last word. The (non-)regulation of the digital economy over many years is a typical negative example in this respect. That is why the EU Commission is now opting for a slow but proactive approach to regulation. Over time, it will only become apparent which applications in individual cases still fall under the term artificial intelligence and in which risk category.
In this way, the EU Commission uses elements of modern tech governance and is moving towards control by the government. Until now, private companies often set the (de facto) rules for the digital space themselves with their products. The legislator enjoyed the role of a fire brigade for particularly serious excesses. As a result, the legislative influence was and is not formative. Now, private parties are to continue to develop innovative models – and to contribute to the specification of the body of standards. However, the legislator is leaving the merely reactive position – anything goes is a thing of the past.
Instead, the current approach fits into the EU’s self-image as a global trendsetter in the digital space. This finding may seem surprising, as digital innovations are not typically associated with the EU. However, it is foreseeable that a pure view of technical innovation falls short. In the coming decade, rule-makers worldwide will try to regain the upper hand in dealing with technology. Whether this will always be to the benefit of fundamental rights is another matter. At least against this background, it may be quietly optimistic that the EU’s approaches will be taken into account in one way or another and potentially have a protective effect in favour of fundamental rights. Copies of European laws are not always necessary (and appropriate) for this. It will often be sufficient for EU legislation to give a convincing reference anchor. In any case, golden years lie ahead for technology regulation.
The blog posts published by bidt reflect the views of the authors; they do not reflect the position of the institute as a whole.
The article first appeared in the Tagesspiegel Background of 15.07.2021.