On the eve of a potential final round of inter-institutional negotiations (‘trilogue’) on legislation on artificial intelligence (‘AI Act’) (see EUROPE 13263/15), the European Consumer Organisation (BEUC) expressed its concern on Monday 23 October that the co-legislators could end up with “a weak and unclear approach” to regulating this technology.
For BEUC, the proposed approach to determining “which types of generative AI systems should comply with which obligations is unclear and complex”. BEUC believes that this approach “would lead to considerable uncertainty for regulators, consumers and businesses”.
In addition, BEUC fears that only generative AI systems developed by very large companies will be adequately regulated, leaving a significant number of other systems to be regulated less rigorously.
BEUC is not alone in its concerns about the outcome of the forthcoming trilogue. In an open letter also dated 23 October, some fifteen associations representing start-ups called on the co-legislators to clear up ambiguities concerning general-purpose AI systems.
The signatories believe that the introduction of provisions on this subject would lead to an environment in which start-ups would not be able to determine whether an AI is risky or not depending on the objective pursued, and that the obligations incumbent on them would be too onerous.
See the letter: https://aeur.eu/f/97j (Original version in French by Thomas Mangin)