The verification of the age of minors and personal data, at the center of the controversy.
The Italian authorities announced on Friday the blocking of the chatbot ChatGPTaccused of not respecting the legislation on personal data and of Lack of age verification system for underage users.
In a statement, the national data protection authority said the decision had “immediate effect” and accused the chatbot of failing to respect those two basic personal data issues.
This decision will result in “the provisional limitation of the processing of Italian user data by OpenAI”, the American society creator of ChatGPThe added.
ChatGPT, launched in November, was quickly adopted by users impressed by its ability to answer difficult questions, write sonnets, and even pass exams.
Asked by AFP, an OpenAI spokesperson assured that the chat was “deactivated for users in Italy” to respect the decision of the data authority.
However, he defended that the company is “concerned about data protection” and considers that it respects the European regulations on the matter.
“We are actively working on reduce personal data in training our AI. We think that an AI regulation is necessary and look forward to working on this with the Authority,” the spokesperson said.
The Italian authority stresses that ChatGPT “suffered on March 20 a data breach (data breach) on user conversations and information relating to the payment of subscribers to the payment service”.
service interruption
After the first notifications of the problem, OpenAI briefly interrupted the service and then recognized an error in an external tool that affected 1.2% of subscribers and an undefined number of free ones.
The surveillance body reproaches “the lack of an informative note for users whose data is collected by OpenAI, but above all the absence of a legal basis that justifies the mass collection and conservation of personal data, in order to train the algorithms that make the platform work”.
In addition, while the robot is intended for people over 13 years of age, “the authority emphasizes that the absence of any filter to verify the age of users exposes minors to answers absolutely not in accordance with their level of development”.
The authority asks OpenAI to “communicate within 20 days the measures adopted” to remedy this situation, “under penalty of a penalty of up to 20 million euros (21.75 million dollars) or up to 4% of annual global turnover”.
This announcement comes after the European police agency Europol warned on Monday that criminals were willing to take advantage of artificial intelligence such as the chatbot ChatGPT to commit fraud and other cybercrime.
ChatGPT is financed by the computer giant Microsoft, which has added it to its services and presents itself as a potential competitor to the Google search engine.
Background in Italy
Italy’s data protection agency also banned the artificial intelligence company Replika from using the personal data of Italian users in early February because it offers to chat with personalized avatars that speak and listen.
Some users had complained about receiving messages and images that were too daring, close to sexual harassment.
The European Regulation of Data Protection (GDPR), which has arranged billions of dollars in fines for the technological giants, could become the biggest enemy in the old continent of the new content-generating AI.
The European Union is preparing a bill to regulate artificial intelligence, which could be ready by the end of 2023 or the beginning of 2024 and be applied in a few years.
Artificial intelligence generates deep fears and not only for the use of personal data.
