Take the opportunity to influence the AI Act

"The regulation has been negotiated and discussed for almost two years. During that time, many have delved into some parts of it, while other parts have not received as much attention”, says Susanne Stenberg, researcher at RISE and legal expert in new technology.

In a so-called position paper, Susanne Stenberg and Håkan Burden have summarized about a year and a half of analysis work on the AI Act. Here, they highlight several parts that they believe have not been so clearly visible in the discussions around the new law proposal. In their analysis, they highlight the emergence of what they call "A Triangle of Trust": three components that together will provide reliable, qualitative, and safe AI systems.

Download and read the analysis

A chance to make an impact

The proposal is now being finalized and decided at the earliest this spring. In parallel with this, technical standards, guidelines, and definitions will be developed to realize the legal requirements. The two European standardization organizations are involved here: the European Committee for Standardization, CEN, and the European Committee for Electrotechnical Standardization, CENELEC.

"The new standards will, among other things, describe what is representative data and relevant data for an AI system. As well as describe what is sufficient reliability and robustness," says Håkan Burden.

The standardization work will continue until the regulation enters around 2025-2026. Here, the two researchers are urging more companies and organizations to get involved. On the one hand, there is a good chance of impact the new standards, and on the other hand, an opportunity to increase one's own knowledge.