Meta has paused its AI expansion plans in Europe due to concerns over using user data from Facebook and Instagram. This decision came after the Irish privacy regulator urged Meta to halt its plans to use data from European users.
According to Reuters, Meta’s delay follows complaints and calls for action from the advocacy group NOYB. This group has pushed data protection authorities in several countries, including Austria, Belgium, France, Germany, Greece, Italy, Ireland, the Netherlands, Norway, Poland, and Spain, to scrutinise Meta’s data practices. The main concern is that Meta’s use of public posts on Facebook and Instagram for AI training may breach EU data usage regulations.
Meta has admitted to using public posts to power its AI models, specifically the Llama models, but insists it does not use restricted updates or private messages. Meta argues that this approach complies with user privacy agreements.
Meta’s response
In a blog post last month, Meta clarified its data usage policies for European users. The company stated, “We use publicly available online and licensed information to train AI at Meta and the information people have shared publicly on Meta’s products and services. This information includes things like public posts or public photos and their captions. In the future, we may also use the information people share when interacting with our generative AI features, like Meta AI, or with a business to develop and improve our AI products. We don’t use the content of your private messages with friends and family to train our AIs.”
Meta has been transparent with European users through in-app alerts about how their data might be used. Despite this, the company’s plans are now on hold until EU regulators review these concerns and their compliance with GDPR.
Impact on creators
This situation presents a challenge because Meta can argue that its data usage is covered under its broad user agreements. However, many users might not realise that their public posts are included in Meta’s AI data pool. For creators aiming to reach a broad audience on Facebook and Instagram, Meta’s AI models could utilise their public content without explicit consent.
When you see AI-generated content from Meta that resembles your work, it likely draws from your public posts. This issue is part of a broader concern regarding AI models and their use of web data. Technically, Meta is correct in saying it has disclosed such practices in its agreements. However, EU officials are likely to demand more explicit permissions, requiring users to agree to their content being used for AI training specifically.
The most probable outcome is that European users must give explicit consent for Meta’s AI models to repurpose their content. For now, this means a further delay in the rollout of Meta’s AI tools in Europe.
