On Tuesday 25 July, the European Consumer Organisation (BEUC) published a set of recommendations for ensuring that the protection of consumers and citizens is a top priority in the interinstitutional negotiations on the AI Act (see EUROPE 13222/3). On a large majority of aspects, BEUC would like the future final agreement to retain the approach of the European Parliament and not that of the EU Council.
In concrete terms, BEUC would like to see the European Parliament’s approach retained on the question of trustworthy AI, the list of high-risk use cases, users’ rights and possible redress, the assessment of the impact on fundamental rights and generative AI.
On this last point, BEUC believes that systems should be subject to “a set of specific rules and not only be regulated when used in a high-risk context”. Persons or entities deploying such systems should monitor and address how the system affects consumers throughout its lifespan.
“Relevant documentation about their risk assessments, including a short and less technical version informing consumers about the potential remaining risks”, should be published and, if a risk is detected, these systems should only be placed on the market if the risk can be mitigated.
BEUC also believes that the co-legislators should return to the original proposal for the classification of high-risk artificial intelligence systems.
See the report: https://aeur.eu/f/8a0 (Original version in French by Thomas Mangin)