The Irish data protection authority (DPC) said on Wednesday 21 May that it was aware of the “significant measures and improvements” made by Meta to the processing and protection of its European users’ public data, which will be used to train its AI models from 27 May (see EUROPE 13629/8).
In its press release, the DPC considers that the improvements introduced by the US firm are in line with the recommendations it made last June when it obtained a temporary suspension of Meta’s training programme (see EUROPE 13433/12).
Meta must also submit a report to the Irish DPC by October 2025 to ensure that the improvements are still in place.
“Our goal at pre-processing stage is to ensure that companies innovate responsibly, mitigate identified harms and risks to individuals and appropriately consider data subjects’ rights by balancing and protecting fundamental rights against the companies’ interests”, says the DPC in its press release.
In December, the European Data Protection Board (EDPB) issued an opinion that did not oppose the use of personal data for training AI models outright, taking the view that the legitimate interest argument could be a legal basis for justification (see EUROPE 13549/10).
But many associations and NGOs are opposed to such use, which they consider to be in conflict with GDPR rules.
See the press release: https://aeur.eu/f/h0d (Original version in French by Isalia Stieffatre)