Meta to train its AI using social media posts from Europe

Meta plans to use social media content posted by European users of its platforms to train its Llama AI models.

The company’s suite of generative AI tools called “AI at Meta” is only available in select countries with Meta planning to roll it out in Europe soon.

In a blog post, Meta said that it would need to train its models using social media content posted by European users of its platforms to make its AI tools more relevant to them. The problem is that European data protection laws are fiercely protective of personal data.

Meta explained, “If we don’t train our models on the public content that Europeans share on our services and others, such as public posts or comments, then models and the AI features they power won’t accurately understand important regional languages, cultures or trending topics on social media.”

The post also pointed out that OpenAI and Google already do this so Europeans should be fine with Meta using their public Facebook and Instagram content to train its models.

The arguments over whether this breaches GDPR privacy laws are ongoing, but Meta’s updated terms of service come into force on 26 June 2024 in Europe.

None of your business

Privacy advocacy group, NOYB (None of Your Business) has filed complaints in 11 European countries, asking the authorities to launch an urgent procedure to stop this change immediately.

NOYB founder Max Schrems said, “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’.”

“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. Meta doesn’t say what it will use the data for, so it could either be a simple chatbot, extremely aggressive personalised advertising or even a killer drone.”

“Meta also says that user data can be made available to any ‘third party’ – which means anyone in the world.”

Schrems said that the logical approach would be for users to opt-in to allow Meta to access their data but Meta is instead going with a “hidden and misleading opt-out form”. At least the Europeans get an opt-out form. That’s not an option for users in other countries.

Meta says it does “not use people’s private messages with friends and family to train our AI systems.”

As for the rest, Meta says it should be fair game unless European users specifically opt out. If Meta goes ahead with this then even old dormant or forgotten European Facebook and Instagram accounts will be added to the AI training data pile.

The post Meta to train its AI using social media posts from Europe appeared first on DailyAI.