Meta: Help Users Stop Accidentally Sharing Private AI Chats

Updated: 3 July 2025

Meta’s AI app includes a “Discover Feed” — a public stream of real AI conversations. But many of these chats appear to have been shared by mistake. From personal medical info to work stress and calendar reminders, people are unintentionally broadcasting private moments.

Why? Because Meta’s app fails to clearly show users that their chats are going public. There’s no clear iconography, no familiar interface cues, and no strong friction to prevent oversharing.

That’s why we’re calling on Meta to fix the app’s design and restore key privacy protections — so users can truly understand when they’re sharing with the world.

More than 10,000 people have already signed this petition, and Meta responded by reaching out to us and briefly implementing a mandatory warning screen to prevent unintentional sharing. Now we need to make that change permanent.

1. Add clear visual indicators to make it unmistakable when a prompt and response are being made public.
2. Clarify sharing through better text prompts and a revised user experience, including at the moment of sharing and anywhere shared content appears. The goal is education and awareness, not just consent buried in unclear buttons.
3. Use familiar platform conventions, like the standard phone share sheet, to ensure users understand they’re posting to a public space.
4. Notify users who have shared content publicly, and provide details about how they can permanently delete any unintentionally shared content.

When privacy is at stake, design matters. Add your name to demand Meta build clear, educational friction into its AI app — and make public sharing a truly informed choice.