Stop in Italy a Chat GPT, the artificial intelligence that is depopulating on the web, until it respects the privacy policy.
And how much asks the Guarantor for the protection of personal data which with immediate effect has ordered the temporary limitation of the processing of data of Italian users against OpenAI, the American company that developed the chatbot. At the same time as the request was open also an'investigation.
Gathering data to train algorithms
In the provision, the Privacy Authority challenges OpenAi with the lack of information to users and all interested parties whose data is collected but above all criticizes thelack of a legal basis justifying the collection. Furthermore, according to the Guarantor “the massive conservation of personal data” primarily serves the “purpose of train algorithms underlying the operation of the platform". The Authority also reproaches the company with the information provided from ChatGPT which – underlines the note – “do not always correspond to the real data, thus determining a treatment of incorrect personal data".
Last March 20 ChatGPT had suffered a data loss (data breach) concerning user conversations and information relating to the payment of subscribers to the paid service.
Absence of systems for verifying the age of minors
Another issue that is disputed is theabsence of age verification systems. According to the terms published by OpenAi, the service is aimed at over 13 years old. According to the Authority missing completely any verification filter of the users' age, a fact that exposes "minors to absolutely unsuitable answers with respect to their degree of development and self-awareness".
The American company, which does not have an office in the Union but has designated a representative in the European Economic Area, will have to communicate within 20 days the measures taken in implementation of what is requested by the Guarantor. The risk is that of one fine up to 20 million of euros or up to 4% of the annual global turnover.
The answer of Artificial Intelligence
Waiting for an official response from OpenAi we asked ChatGPT directly if “is it true that it collects private data illegally? There Artificial Intelligence response was: "No, it's not true that I collect private data illegally. As a virtual assistant, I am designed to respect the privacy of my users and do not have the ability to collect or store personal information or private data unless strictly necessary to perform my function as a virtual assistant. Also, as a language model of OpenAI, my creation and operation are subject to stringent security and privacy controls. OpenAI is committed to complying with privacy and data security regulations and adopting cybersecurity best practices. If you have concerns or concerns about data collection or data security on the Internet, I recommend that you consult your country's privacy and data security guidelines and follow best practices to protect your privacy online".
