AI governance decisions by European Parliament: two steps forward, one step back
By Prof. dr. Evert F. Stamhuis LLM & Prof. dr. Klaus Heine
Erasmus School of Law
Erasmus Centre for Data Analytics
Jean Monnet Centre of Excellence Digital Governance
Artificial Intelligence was the central topic in the three proposals that the European Parliament (EP) adopted on Oct. 20. Three Parliament resolutions set the stage for the development and deployment of artificial intelligence, robotics and related technologies in the EU. The resolutions comprise a Framework for the ethical aspects of artificial intelligence, robotics and related technologies, a Civil Liability regime for artificial intelligence and a Resolution on the intellectual property rights for the development of artificial intelligence technologies. https://www.europarl.europa.eu/doceo/document/TA-9-2020-10-20-TOC_EN.html
In this position statement we want to share our first assessment of these policy documents in the light of our research on governance of digitalization today and tomorrow (www.digov.eu). In sum, we hold these decisions to be an important move forward for the EU to show where we stand in the global playing field of breakthrough technologies. A clear policy helps to avoid unhealthy competition between member states and to prevent a race to the bottom for the involved industries. However, the instrumental approach of the European Parliament is in our view disappointingly conservative and insufficiently futureproof.
On the ethical framework we will stress that the risk-based approach and the principle of human centric technology application are a good starting point. The elaboration however does not end well. On the liability issue we will point at the important policy choice in favor of the operators’ liability, which is promising under certain conditions. The intellectual property policy resolution we appreciate as an example of running without moving an inch.