Italy to Lift ChatGPT Ban if OpenAI Complies with EU Regulations
Italy’s data privacy regulator has announced that it plans to lift the ban on OpenAI’s ChatGPT service, which has been in place since it was introduced, by the end of April. This is subject to OpenAI making certain changes to bring the service into compliance with EU privacy regulations.
Should OpenAI agree to meet specific terms, April 30 is the earliest date the service could be operational again. However, the regulator has stipulated that compliance must be achieved by September 30 at the latest. Failure to meet the required conditions will result in a more permanent ban.
The measures that OpenAI must implement in order to get the ChatGPT ban lifted include:
- Age-gating the service to ensure that users are at least 18 years of age
- Collecting explicit consent from users to add their data to its training models
- Allowing anyone, whether they used the service or not, to check for false personal information and to request its deletion
- Posting a privacy notice on its website explaining how it uses and stores collected data
- Running an advertising campaign across Italy on “radio, television, newspapers and the web” to inform the public of all changes
The ban was imposed by the Guarantor for the Protection of Personal Data (GPDP) due to privacy concerns. The GPDP was troubled by transparency, data collection, and use practices, as well as the absence of any oversight for minors who entered sensitive personal information without adequate protection. The ChatGPT service will need to demonstrate greater transparency in order to regain access to the Italian market.
OpenAI will also need to age-gate the service so as to prohibit children under 13 from using it and require that those aged between 13 and 18 obtain a parent or guardian’s consent upon creating an account. Although an “age verification process” is called for, it is not clear what this will entail beyond self-certifying one's age. OpenAI is presently “looking into” options in this area.
Italy is currently the only country in western Europe to have imposed a ChatGPT ban. However, a similar ban has been considered by Germany, which has approached Italy for advice. ChatGPT is now under investigation by France’s lead data protection agency following several privacy-related complaints. Meanwhile, Spain has asked the European Data Protection Board to bring up the topic of chatbot regulation with other European nations with a view to developing unified standards.
However, regulation of ChatGPT does not yet appear on the cards in most countries. The only nations that have implemented bans on ChatGPT are those that OpenAI prohibits from using it, such as China, North Korea, and Russia. Several of these countries have gone ahead with their own comparable large language models, highlighting fears of an “AI arms race”. While many governments are concerned about privacy and safety, some worry about falling behind in the AI race and the impact on private sector productivity and innovation.
Despite these considerations, regulations around ChatGPT and other similar AI technologies seem increasingly inevitable given the potential risks which have already been identified. There are also concerns about their impact on education, as students attempt to use AI for homework and tests.
Dr. Ilia Kolochenko, Founder of ImmuniWeb, believes that ChatGPT may soon face regulatory troubles in the same way that GDPR did: “Many countries are actively working on new legislation for all kinds of AI technologies, aiming at ensuring non-discrimination, explainability, transparency and fairness – whatever these inspiring words may mean in a specific context, such as healthcare, insurance or employment. Of note, the regulatory trend is not a prerogative of European regulators.”
However, banning AI is not seen as a viable solution as law-abiding companies will simply follow the ban while hostile nations and threat actors will continue their research and development, gaining unfair advantage in the global AI race.
Italy has already banned the chatbot Replika.ai due to similar concerns about the safety of minors. In 2019, Facebook had to ban a chatbot operated by Israel Prime Minister Benjamin Netanyahu’s page twice for sharing polling information illegally and for promoting conspiracy theories about Arab residents.
