LONDON (AP) — European Union regulators said Thursday they are investigating one of Google’s artificial intelligence models over concerns about its compatibility with the bloc’s strict data protection rules.
Ireland’s Data Protection Commission has launched an investigation into Google’s Pathways Language Model 2 (also known as PaLM2). The investigation is part of a broader effort, including by other national regulators across the 27 countries, to examine how AI systems handle personal data.
Google’s European headquarters are in Dublin, so the Irish regulator acts as the company’s lead regulator in relation to the bloc’s privacy framework, the so-called General Data Protection Regulation (GDPR).
The Commission said its investigation aimed to determine whether Google had assessed whether data processing by PaLM2 was likely to result in a “high risk to the rights and freedoms of individuals” in the EU.
Large language models like PaLM2 are massive collections of data that serve as building blocks for artificial intelligence systems. Google uses PaLM2 to power a number of generative AI services, including email summarization. The company did not respond to a request for comment.
The Irish regulator said earlier this month that Elon Musk’s social media platform X has agreed to permanently stop processing user data for its AI chatbot Grok. The platform only did so after the regulator took it to court the month before, filing an emergency application in the High Court to get X to “suspend, restrict or prohibit” the processing of personal data in its users’ public posts.
Meta Platforms has put on hold its plans to use content from European users to train the latest version of its large language model following pressure from Irish regulators. The decision was “the result of intensive discussions” between the two, the regulator said in June.
The Italian data protection authority temporarily banned ChatGPT last year due to data breaches and required chatbot maker OpenAI to meet a series of demands to address its concerns.