Unlock Editor's Digest free of charge
FT editor Roula Khalaf selects her favourite stories on this weekly newsletter.
Google is currently being investigated by Europe's data protection authority over its processing of private data in developing one among its artificial intelligence models, as regulators increasingly scrutinize the AI ambitions of major tech firms.
The Irish Data Protection Commission, which is accountable for enforcing the European Union's General Data Protection Regulation (GDPR), said it had launched a statutory investigation into the tech giant's Pathways Language Model 2 (PaLM 2).
Launched in May 2023, PaLM 2 predates Google's latest Gemini models that power the corporate's AI products. Gemini, launched in December of the identical 12 months, is now the core model behind the text and image generation offering.
The investigation goals to find out whether the corporate has breached its obligations under the GDPR with regard to the processing of private data of residents of the EU and the European Economic Area.
Under this framework, firms must perform an information protection impact assessment before processing such information if the way through which it’s used is prone to present a high risk to the rights and freedoms of natural individuals.
This is especially true for brand new technologies and is “crucial to making sure that the elemental rights and freedoms of people are adequately taken into consideration and guarded,” the regulator said in a press release.
The assessment is being reviewed as a part of the investigation. Google didn’t immediately reply to a request for comment.
This is the newest in a series of actions by the DPC against the massive technology firms that develop large language models.
In June, Meta paused plans to coach its Llama model on public content shared by adults on Facebook and Instagram across Europe following talks with the Irish regulator. Meta then restricted the supply of a few of its AI products to users within the region.
A month later, X users discovered that they’d chosen to make use of their posts on the positioning to coach systems at Elon Musk's xAI startup.
The platform suspended the processing of European user data to coach its Grok AI model in August after the DPC took legal motion, the primary time the regulator used its powers to take such motion against a technology company.
