Opinions on the Proposed European Artificial Intelligence Act
Publication: ZRVP
The European Data Protection Supervisor (EDPS) and the European Data Protection Board (EDPB) have recently released a joint opinion (the Opinion) on the proposal for a Regulation laying down harmonized rules on artificial intelligence (AI Act) which aims at ensuring the ultimate purpose of putting forward legislation for a coordinated European approach on the human and ethical implications of AI.
While the Opinion is not binding on the European Commission, the EDPS and EDPB are directly responsible for the enforcement of EU and national Data Protection rules. So, their views are likely to be influential with the European institutions.
Although the two European agencies have not challenged the advantages the legislative proposal brings, they do highlight a few key aspects and consequences which should be considered and even modified by the AI Act as it is drafted at this moment.
First, the two agencies call for a general ban on any use of AI for automated recognition of human features in publicly accessible spaces. Undetectable human characteristics should cover not only face recognition, but also other personal features such as voice, fingerprints or DNA which should not be detected by AI applications. In the same manner, it is brought to light the need to avoid any kind of social scoring through such technology which would basically risk affecting some of the most important human rights. Creating the context for discrimination based on grounds like ethnicity, gender, political or sexual orientation is harshly criticized as it could even risk affecting human dignity.
In the same manner, the Opinion states that detecting specific human emotions should only be applicable in very rare cases, thoroughly described and controlled, such as for detecting health issues.
On a more applied approach, it has also been stressed out the need to set up specific national authorities which would essentially be in charge with supervising and enforcing the AI Act. Even if many advantages are brought by such new national bodies, however, a high degree of autonomy should be ensured, which should be considered also for the European Artificial Intelligence Board (EAIB). In this way it ought to be ensured that their independence is not undermined by any external political influence, and it should contribute to a harmonized regulatory approach and to a consistent interpretation of the legal provisions across all EU.
Another important aspect criticized by the two authorities refers to the lack of any specific compulsory marks for the AI systems which would set a degree of safety when using such technology. Therefore, the Opinion stresses that the AI Act should be better aligned with data protection rules.
Despite the many advantages brought by artificial intelligence, the Opinion raises the alarm before the European Commission by underlying some aspects which should be improved to avoid as much as possible bringing about new risks or negative consequences for individuals or the society.