The president of the data protection authority spoke about the risks of artificial intelligence.
Attila Péterfalvi emphasized that the new technology contains many more risks due to the much larger amount of data, and due to its operation, it is difficult for ordinary people to understand.
The EU Regulation on Artificial Intelligence (AI) entered into force on August 1st, but the regulation is more important from the point of view of consumer protection and product liability, and does not cause much change in the short term in data protection matters. Among other things, because the GDPR (general data protection regulation), which came into force in 2018, already regulates several issues more strictly than this regulation, Attila Péterfalvi, the president of the National Data Protection and Information Freedom Authority (NAIH), told our newspaper.
The NAIH would definitely play a role
"However, in the longer term, according to the recently adopted statement of the European Data Protection Board, the supervisory role of the data protection authorities according to the MI regulation should also be considered. This idea is also supported by the Hungarian authorities"
Péterfalvi emphasized. According to his information, the EU regulation enables the testing of AI in a regulatory test environment (sandbox) and under real conditions under certain warranty conditions.
According to the regulation, the sandboxes must be operated by a supervisory authority to be designated by the member states, however, if it is not the NAIH, the data protection authority must also participate in the case of personal data based on the MI regulation.
Since the use of AI in many cases involves the handling of huge amounts of personal data, and also during the performance of tasks according to the GDPR, there may be a need to test the operation of the AI algorithm,
NAIH is currently working on the creation of a new MI lab in addition to its forensic lab that has been operating for a few years.
Even basic rules were not followed
Péterfalvi explained: AI has spread exponentially in many areas in recent years. On the one hand, this provides many opportunities for more efficient operation, which, for example, is used by the Hungarian authorities to automatically transcribe audio recordings, however
due to the much larger amount of data, AI also harbors many more risks, and due to its operation, it is difficult for ordinary people to understand.
For this reason, compliance with the GDPR requires a more serious effort on the part of the data controller compared to an average data management.
The entire article can be read on Mandine!
Featured image: Gerd Altmann / Pixabay
