The tech giant says that it’s “disappointed” by the regulatory request and that it represents “a step backwards for European innovation”.
However, the Irish request was made following unease and uncertainty over how Meta would train its new AI systems using Meta users’ personal data from photos, comments and other content on public posts.
The company will now temporarily shelve its EU-based AI data collection process, which means that its AI services here will also now take a hit.
Meta’s AI aims to analyse and record content from adult users to help create a ‘large language model’ (LLM), used as the basis for developing AI responses to user queries and prompts. In recent weeks, it has been notifying users across Europe that it would collect their data, offering an opt-out. But it still fell short of satisfying European GDPR and privacy rules.
“The DPC welcomes the decision by Meta to pause its plans to train its large language model using public content shared by adults on Facebook and Instagram across the EU and EEA,” said the DPC in a statement.
Today’s News in 90 seconds – 14th June 2024
”This decision followed intensive engagement between the DPC and Meta. The DPC, in co-operation with its fellow EU data protection authorities, will continue to engage with Meta on this issue.”
In a statement, Meta called the move “a step backwards for European innovation” and said that it still hoped to launch its AI data services in Europe.
The regulatory decision is a blow to the tech giant, which employs around 3,000 people in Dublin and competes with OpenAI, Google, Anthropic and other large AI-focused companies for a slice of tech’s boom sector.
Meta already has a large language model called Llama, which is on its third version. Its AI data collection for Llama has already been developed outside the EU.
The Meta move may also focus minds at Google and OpenAI — which owns ChatGPT — both of which currently collect personal data in the EU to train their models.
Meta’s AI data collection plans were recently the subject of a complaint from Max Schrems, the privacy campaigner and long-time Meta adversary. In an intervention request to multiple European regulators, he said that Meta’s legal basis for collecting personal data needed to be challenged.
“Meta is basically saying that it can use any data from any source for any purpose and make it available to anyone in the world, as long as it’s done via AI technology,” he said.
”This is clearly the opposite of GDPR compliance. ‘AI technology’ is an extremely broad term. Much like ‘using your data in databases’, it has no real legal limit.”
However, Meta described the Irish regulatory request a blow to innovation in Europe.
“We’re disappointed by the request from the Irish Data Protection Commission, our lead regulator, on behalf of the European DPAs, to delay training our large language models using public content shared by adults on Facebook and Instagram, particularly since we incorporated regulatory feedback and the European DPAs have been informed since March,” the company said in a response.
“This is a step backwards for European innovation, competition in Al development and further delays bringing the benefits of Al to people in Europe.
“We remain highly confident that our approach complies with European laws and regulations. Al training is not unique to our services, and we’re more transparent than many of our industry counterparts.
“We are committed to bringing Meta Al, along with the models that power it, to more people around the world, including in Europe. But, put simply, without including local information we’d only be able to offer people a second-rate experience. This means we aren’t able to launch Meta Al in Europe at the moment.
We will continue to work collaboratively with the DPC so that people in Europe have access to, and are properly served by, the same level of Al innovation as the rest of the world.”
The company added that the delay would also enable it “to address specific requests we have received from the Information Commissioner’s Office, our UK regulator, ahead of starting the training”.
Separately, the DPC also confirmed to the Irish Independent that it had been contacted by Apple regarding its ‘Apple Intelligence’ AI plans unveiled this week at the tech giant’s Worldwide Developer Conference (WWDC).
“Apple informed the DPC of the announcements and anticipated timelines for delivery of these products,” a spokesperson said.
“Apple also confirmed that it will engage with the DPC more substantively on specific products at the appropriate time, at which point the DPC will undertake the appropriate assessments. In the meantime the DPC continues to monitor developments.”