Based on the latest developments, Germany is expected to approve the current draft of the AI Act this coming Friday. eco – Association of the Internet Industry advocates for a real-world transposition and EU-wide standardised criteria in dealing with AI systems.
“The mistakes of the GDPR must not be repeated – Germany or other EU Member States should not go down special national paths in regulating GPAI models and high-risk systems,” says eco Chair Oliver Süme. “For a true level playing field in Europe, there is a need for harmonised obligations, requirements and standards.”
According to Süme, this requires close and regular exchanges between all stakeholders. This is particularly important in order to remain flexible in the event of new use cases and to develop practical criteria for assessing risks. Furthermore, it is currently unclear as to what measures companies must take to minimise these risks. “There is an urgent need for the EU to make improvements on this and to provide clarity,” says Süme. “Otherwise, the AI Act will become a brake on AI-driven innovation throughout Europe and distort international competitiveness.”
At the same time, from the perspective of the association, the German federal government should take advantage of possible leeways in the AI Act with regard to real-time biometric surveillance. Süme states: “For AI to be successful, it needs to be accepted by the public. However, the provisions in the AI Act for real-time biometric surveillance increase concerns about jeopardising civil rights and could completely undermine trust in artificial intelligence. The German federal government had ruled out real-time biometric surveillance in its coalition agreement and, to avoid losing public trust, it should remain consistent in this regard.”