Facebook Uses User Data to Train AI: Is Opting Out Effective?

Meta, the parent company of Facebook, is advancing its generative AI tools. Now, they offer an option for users to ask for the removal of certain personal data used for AI training, but there are many conditions attached.

Earlier in the year, Mark Zuckerberg, the CEO, unveiled plans to incorporate AI functionalities into Meta’s platforms including Facebook, Instagram, and WhatsApp. Generative AI, while popular in Silicon Valley, has raised concerns due to its unclear legal status and rapid progression.

Want to prevent Meta from utilizing your data for AI enhancement? They've introduced a form in their help center titled “Generative AI Data Subject Rights at Facebook.” This form allows users to ask Meta for access to third-party data used in AI and to request deletion of personal details. However, "request" is the key term. There's no assurance that the company will delete the data or even share the requested information.

This form doesn't relate to the vast amount of user information already gathered by Meta on its platforms. It's specifically about external data sources Meta might use to enhance its AI, like online information or data bought from third-party brokers.

To submit the opt-out request, users don't need a Facebook account. Only their country, name, and email are required. From personal experience, using the desktop site for this form was smoother than the mobile version, and it took over a day for Meta to confirm the request.

For US users, the efficacy of this form remains uncertain. However, UK data protection laws might make this more beneficial for UK residents. Thomas Richards, a Meta representative, informed Gizmodo that the rights users can exercise regarding their data might differ based on their location. However, Meta didn’t provide any comments to WIRED despite multiple attempts.

A company post shared a bit more about how Meta uses personal data for AI training. It mentions, “In the European region and the UK, we base our collection and processing of personal data on legitimate interests. In other applicable regions, we ensure there's a legal foundation.”

This initiative by Meta follows a $1.3 billion penalty they faced from European regulators for mishandling UK-originated data.