Since the global launch of China’s new artificial intelligence company DeepSeek on 20 January, four major national authorities have already looked into the problems raised by the model and its data collection.
Firstly, Italy, which expressed serious doubts early last week about DeepSeek’s compliance with the GDPR, citing “potentially high risks of infringement of the European regulation”.
The national regulator wants to “clarify the learning methods used by AI”. Since Thursday 30 the application has no longer been accessible or downloadable in Italy, and the authorities have announced that they are opening an investigation.
Ireland and Belgium followed suit. On Wednesday 29 January, Dublin requested “information on data processing carried out in relation to data subjects in Ireland”.
The Belgian Data Protection Authority (DPA) announced the following day that it was opening an “investigation” following a complaint.
France has now joined the fray, sending a request for information to the company on Thursday 30 January. As with the other three regulators, the French CNIL is concerned about how the application works and whether it respects users’ privacy, and has announced that it is “analysing” the tool.
The EU says that it is taking note and closely monitoring the actions of the Member States, and points out that DeepSeek, in its capacity as a “hosting service”, will have to comply with the general obligations of the regulation on digital services in the coming months. (Original version in French by Isalia Stieffatre)