Meta keeps launching AI features that assume consent: Why it keeps backfiring
Essential brief
Meta recently launched an Instagram feature allowing AI-generated images from public accounts without prior consent, prompting swift backlash and removal. This marks the fourth instance in under tw
Key topics
Key facts
Highlights
Why it matters
Meta's repeated reliance on silent consent for AI data use underscores ongoing challenges in balancing AI innovation with user privacy rights. Regulatory pushback in multiple regions highlights the growing demand for transparent and consent-based AI practices. How Meta adapts its consent policies will influence broader industry standards and user trust in AI technologies.
Meta introduced a new Instagram feature last week that enabled anyone to create AI-generated images using photos from public accounts without obtaining prior permission. The feature faced immediate criticism, leading Meta to disable it within days. This incident is the fourth time in less than two years that Meta has launched AI-related products based on the assumption of silent consent from users.
In 2024, Meta used public content from Facebook and Instagram profiles across the EU to train its AI models, automatically opting users in and requiring them to submit an objection form to opt out. Privacy group NOYB challenged this practice, arguing that the EU's highest court ruled that "legitimate interest" cannot justify violating data protection rights, whether for advertising or AI training. Complaints were filed in eleven EU member states, revealing that opting out required users to log in, complicating the process.
Brazil took a stricter stance, with its National Data Authority ordering Meta to stop using Brazilian data for AI training due to risks of irreparable harm. Unlike in Europe, Brazilian users did not receive notifications, and the opt-out process was buried within Meta's Privacy Policy Center, making it difficult to navigate.
By 2025, the debate shifted from whether Meta informed users to whether an opt-out policy was legally acceptable at all. Meta continued to rely on the "legitimate interest" opt-out model, adding that opting out after AI training had started would not remove previously collected data, conflicting with the right to be forgotten.
The recent Instagram "Muse AI" feature followed the same pattern: public accounts were opted in by default without notification, leading to viral backlash and a quick reversal. Meta's typical response is to roll back features only after regulatory pressure or public outcry, rather than adopting opt-in consent models proactively.
This recurring approach reflects Meta's prioritization of maintaining large data sets for AI development, as opting in reduces available data and potentially impacts product quality. While this gamble often succeeds, recent events show it can backfire, raising questions about when the financial or reputational costs will force Meta to change its consent strategy.
Key topics in this update include meta keeps launching ai features that assume consent, meta keeps launching, and assume consent.