The Italian privacy regulator has ordered ChatGPT to be banned effective from its chatbot. It accuses OpenAI creators of illegally collecting personal data and ordered them to cease immediately.
The press release by the country’s Data Protection Authority (the GPDP) this morning stated that OpenAI lacks legal justification to collect personal data. OpenAI, according to the GPDP, has no safeguards in place to prevent underage users from accessing its service. This “exposes minors [to absolutely inappropriate answers] in relation to their level of development and self-awareness.
It is not the first time Italy’s data watchdog has taken this action against an AI chatbot. In February , Replika.ai was banned by the regulator. Replika is well-known for the close relationships that users have with their chatbot. Many expressed distress after Replika removed the option to engage in erotic roleplay.
The Verge still had access to ChatGPT at the time of publication. This was via a VPN that redirects traffic through Milan, Italy. OpenAI will need to take time to comply with the Italian data protection order. According to the GPDP, OpenAI has twenty days to inform regulators about the measures taken and possible fines up to EUR20 millions (or 4 percent of its annual global turnover).
According to the authority, ChatGPT lacks the legal basis for mass collecting and storing personal data in order to “train” its algorithms. It also stated that the company processes data incorrectly.
ChatGPT was also affected by a data breach that exposed chats and payment information last week, according to the Italian authority. OpenAI doesn’t verify users’ age and exposes minors to “absolutely unsuitable answers” in relation to their level of self-awareness and development.
OpenAI does not have a representative in the EU, but it has 20 days to inform the European Economic Area about how it intends to bring ChatGPT in compliance with EU privacy rules. If it fails to do so, the company could face a fine of up to 4% of its global revenues.