OpenAI, the producer of ChatGPT, has announced plans to address concerns raised by Italy's data protection agency Garante, following a temporary ban on the chatbot. The move came after Garante accused OpenAI of failing to verify the ages of ChatGPT users and of gathering personal data without a legal basis.
In response, OpenAI pledged to be more transparent in how it handles user data and to provide measures to address Garante's concerns. In a blog post titled Our Approach to AI Safety, OpenAI stated that it does not use data for selling services, advertising or building profiles of individuals. The company's goal is to develop nuanced policies against behaviour that represent real risk to people. It removes personally identifiable information from its datasets where possible and fine-tunes its models to reject user prompts asking for such information.
Sam Altman, OpenAI's CEO, attended a video conference with Garante on Wednesday to talk about the measures being taken. The company plans to send a document detailing the measures to Garante on Thursday. Garante will assess the proposals made by OpenAI but warns that it will not hesitate to take further action if the measures are insufficient.
The ban has been sparked by interest from other European privacy regulators who are studying whether to coordinate similar actions against chatbots and whether harsher measures are needed to protect users.
This is not the first time that Garante has taken action against AI chatbots. In February, it banned Replika, an AI chatbot company, from using the personal data of Italian users, citing risks to minors and emotionally fragile people. It remains to be seen whether other countries will follow Italy's lead in regulating chatbots and other AI systems.