Italy’s data protection authority (Garante per la protezione dei dati personali) has decided to block access to the ChatGPT neural network chatbot in the country “until it complies with privacy rules”, according to a statement. message on the department’s website. This is the first case in the world of this neural network being blocked based on privacy laws, Remarks Cable.
FDA documentation states that ChatGPT, developed by US startup OpenAi, does not comply with European data protection regulations (Gdpr). The decision to block access to the chatbot, the Italian department links to the leak of personal data of ChatGPT users, including became known March 20.
Management spokesperson Guido Scorza explained that OpenAi uses users’ personal data to train artificial intelligence.
“Anyone involved in medical research must obtain consent to conduct experiments. Even those experimenting with new technologies should make the process transparent,” the official notes.
The management stated that OpenAi never provided information on the processing of user and interested party data. In this case, “there is no legal basis that justifies the massive collection and storage of personal data with the aim of “training” the algorithms that underpin the operation of the platform”, specify the experts of the department.
Moreover, they pointed out a problem with the age limit: although ChatGPT is intended for users over 13 years old, its interface does not have age control functionality, and the chatbot “often gives minors totally inappropriate responses to their level of development”. and self-awareness.
Since OpenAi does not have an office in the European Union, the documents containing the Italian regulator’s requirements were sent to the company’s representative in Ireland. The agency expects a response within 20 days: the American company must submit an action plan to eliminate the violations, failing which it risks a heavy fine – 4% of its annual turnover (about 20 million euros). euros). Similar requirements are sent to the United States, to the headquarters of a startup.
Guido Scorza explained to Wired that blocking a chatbot is a temporary measure, which, moreover, can be circumvented using known methods. However, according to him, the management in this matter above all wants to receive information from OpenAi on how it collected user data, whether it received consent for its collection and processing – and whether it complied the conditions if the user refused.
At the end of April, there will be a meeting of the representatives of all the public privacy protection authorities of the countries of the European Union, and the decision of the Italian government will be discussed there. “Nobody wants to hold back innovation, but you can’t develop new technologies at the expense of people’s rights,” Scorza concluded.
ChatGPT commented on the Italian authorities’ decision at the request of RTVI, citing the protection of personal data as an important aspect of the development and use of the technology.
“If Italy considers that the use of ChatGPT violates its personal data protection laws, then it may take measures to restrict or prohibit the use of this technology in its country. after a thorough examination of the legislation and evidence of a violation of the law. However, this decision may be controversial and its consequences may be challenged in court,” the chatbot warned.
Asked about the risks of the service being blocked in other countries, ChatGPT replies that such a scenario is indeed possible – along with the adoption of new laws and rules that specifically regulate the use of neural network services.
Among the probable reasons for the blocking in different countries, ChatGPT, in addition to the processing of personal data, cited the use of a chatbot to create counterfeit and malicious content, phishing, espionage and “contravention of government policies », the creation and distribution of malicious software.
