Since its launch at the end of November, many have already tested ChatGPT, this artificial intelligence capable of answering almost all questions. Some, “to see”, and without being too suspicious, entrusted him with personal information, others even shared industrial secrets. But what does ChatGPT do with our data? Governments are beginning to take notice.
Article written by
Posted
Update
Reading time : 2 min.
/2023/04/09/643284a2ea2f7_gettyimages-1250799402.jpg)
Italy led the way on March 31, banning ChatGPT for multiple privacy and under-18 data breaches. Immediately, several other European countries got in touch with the Italian regulators, in particular Ireland and France, where the CNIL has therefore just registered two complaints, but also Germany, which could take such a radical decision. than Italy in the coming weeks, citing security and privacy concerns.
An imminent European regulation on AI
Meanwhile, the European Parliament – which has been working on the subject for two years – plans to vote, by the end of the month, the proposal for a regulation that frames artificial intelligence in Europe. But in Canada too, ChatGPT is now the subject of a very official investigation by the Privacy Commissioner. An unfavorable opinion could be the starting point for a real legal procedure with, potentially, a ban on ChatGPT also in Canada.
It must be said that the Samsung “case” has been an opportunity for a few days to raise awareness. Three Samsung engineers have indeed wanted to simplify their lives. They therefore entrusted ChatGPT with the correction of errors in computer code, and also the improvement of processes for identifying defective processors of the Korean group: three data leaks in three weeks have been identified, not to mention the preparatory documents for a meeting, entrusted to ChatGPT, to transform them into Powerpoint type presentations. Result: Samsung is considering banning ChatGPT and has launched the development of its own equivalent for internal use.
Two first complaints in France
In France, the first two complaints were therefore addressed to the CNIL, which was already providing advice, in 2021, for the respect of the rights of individuals in the face of chatbots (conversational robots). The first complaint dates from April 4, as revealed by the media The informed.
Lawyer Zoé Vilain criticizes OpenAI, the company behind ChatGPT, for failing to comply with the GDPR, the European General Data Protection Regulation, by not responding positively to her request for access to its own personal data. The second complaint comes from a developer already known for opposing CCTV.
So, is there a risk of a ban in France? For Jean-Noël Barrot, the Minister Delegate for Digital in The gallerya ban “is a wrong answer”. But no answer would be a bad answer too, when you know that 2.6% of ChatGPT users have already revealed confidential information to it, according to a study regularly updated by Cyberhaven.
View comments