Meta Resumes E.U. AI Training Using Public User Data After Regulator Approval

Avatar
Meta has announced that it will begin to train its artificial intelligence (AI) models using public data shared by adults across its platforms in the European Union, nearly a year after it paused its efforts due to data protection concerns from Irish regulators. “This training will better support millions of people and businesses in Europe, by teaching our generative AI models to better

Meta has announced that it will begin to train its artificial intelligence (AI) models using public data shared by adults across its platforms in the European Union, nearly a year after it paused its efforts due to data protection concerns from Irish regulators.

“This training will better support millions of people and businesses in Europe, by teaching our generative AI models to better understand and reflect their cultures, languages, and history,” the company said.

To that end, users’ posts and comments, as well as their interactions with Meta AI, are expected to be used for training and improving the models. It does not cover private messages sent between friends and family and data from accounts below the age of 18.

Users in the region will start receiving notifications this week, both in the apps and via email, that detail the kinds of data the company will be using for this purpose and why it matters in the context of improving AI and the overall user experience.

The notifications will also include an opt-out link to a form where users can choose to object to their public data being collected for AI training. Meta said it will honor all objection forms it has already received, along with newly submitted ones.

The development comes shortly after the European Data Protection Board (EDPB) approved the rollout owing to it meeting legal obligations under the bloc’s stringent data protection laws. Last month, the social media giant launched Meta AI in the European Union across its digital real estate: Facebook, Instagram, WhatsApp, and Messenger.

The company has also pointed out that it’s “following the example set by others including Google and OpenAI,” adding both companies used data from European users to train and fine-tune their respective models.

News of Meta’s AI training plans coincides with a report from Apple about how it uses techniques like differential privacy and synthetic data generation to improve several features like Genmoji and Image Playground, Image Wand, Memories Creation, and Writing Tools in Apple Intelligence without sacrificing user privacy when they choose to send Device Analytics information to the company.

Found this article interesting? Follow us on Twitter and LinkedIn to read more exclusive content we post.

 The Hacker News 

Total
0
Shares
Previous Post

Taiwan charges Chinese ship captain with breaking subsea cables

Next Post

Gladinet’s Triofox and CentreStack Under Active Exploitation via Critical RCE Vulnerability

Related Posts

Explosive Growth of Non-Human Identities Creating Massive Security Blind Spots

GitGuardian's State of Secrets Sprawl report for 2025 reveals the alarming scale of secrets exposure in modern software environments. Driving this is the rapid growth of non-human identities (NHIs), which have been outnumbering human users for years. We need to get ahead of it and prepare security measures and governance for these machine identities as they continue to be deployed, creating an
Avatar
Read More

Ballista Botnet Exploits Unpatched TP-Link Vulnerability, Targets Over 6,000 Devices

Unpatched TP-Link Archer routers have become the target of a new botnet campaign dubbed Ballista, according to new findings from the Cato CTRL team. "The botnet exploits a remote code execution (RCE) vulnerability in TP-Link Archer routers (CVE-2023-1389) to spread itself automatically over the Internet," security researchers Ofek Vardi and Matan Mittelman said in a technical report shared with
Avatar
Read More

CERT-UA Warns of Cyber Scams Using Fake AnyDesk Requests for Fraudulent Security Audits

The Computer Emergency Response Team of Ukraine (CERT-UA) is warning of ongoing attempts by unknown threat actors to impersonate the cybersecurity agency by sending AnyDesk connection requests. The AnyDesk requests claim to be for conducting an audit to assess the "level of security," CERT-UA added, cautioning organizations to be on the lookout for such social engineering attempts that seek to
Avatar
Read More