Italy Gives OpenAI a Chance to Prove its Privacy

Italy became the first European country to ban ChatGPT for its citizens. Now there might be good news for OpenAI. Italy has provided the company with a set of demands if it wants to resume its services in the country. 

The Italian data protection authority has outlined the requirements that OpenAI will have to fulfil before the end of the month to lift the ban on ChatGPT. The watchdog known as Garante has revealed in a press release the measures that OpenAI has to take to get its services back.

This comes after OpenAI accepted the Italian government’s demand of rolling back its services from the company. Sam Altman had tweeted about how he believes that the company is following all the privacy laws of the country, and thus looks forward to starting working again. 

The Italian authority states that OpenAI is required to publish details on its website explaining the procedures and purposes for handling the personal data of both its users and non-users, and to allow individuals to modify or remove their data, complying with the “right to be forgotten.” This should be done by providing easy-to-access tools for the people.

The company must also immediately adopt age gating measures to prevent minors from accessing ChatGPT, adopting more age verification measures. 

Moreover, according to the regulatory authority, OpenAI may only use personal data to train ChatGPT's algorithms if it obtains consent or has a "legitimate interest.” The company should also clarify the legal basis for processing people’s data. 

Italian regulators expressed reservations about OpenAI's ability to collect substantial amounts of data to educate ChatGPT's algorithms and the potential for the system to generate erroneous data about individuals, questioning whether there was a lawful basis for the company's actions.

Italy's regulatory body has mandated that OpenAI conduct an advertising initiative by May 15, using radio and television, newspapers, and the internet, to educate the public about how it employs their personal data to train algorithms.

The European Union’s General Data Protection Regulation (GDPR) is applicable everywhere personal data is being processed. And that is definitely true in the case of LLM based chatbots like ChatGPT. Italy’s ban has also given the EU the chance to build more privacy regulations, and countries like Germany, France, and Ireland are also speaking with Italy to understand more about the reasons behind banning the chatbot.

ChatGPT has landed OpenAI in legal troubles in several countries because of its privacy and security concerns. Apart from Italy, Canada is also investigating OpenAI against a complaint filed for collection, use, and disclosure of personal information without consent.

OpenAI has been working on its safety policy and is in a bid to build “increasingly safe AI systems”. But even then, the company hasn’t taken any concrete steps about it.

Recently, Greg Brockman also made a big tweet explaining how OpenAI is speaking to different governments to comply with policies and also build a more aligned AI model.