Your Meta AI chats might be public, and it’s not a bug
- Abeera Marium Siddiqui
- June 17, 2025
- 10:43 am
- 38
- Technology
The Meta AI privacy leak isn’t just a glitch—it’s a real issue. Many users are unknowingly sharing their private chats on the internet. These conversations, often filled with personal details, are showing up on Meta’s public Discover feed. And no, it’s not a bug. It’s how the system is designed.
Meta AI now works across several apps, including Facebook, Instagram, WhatsApp, and its own standalone app. With over 1 billion users, Meta wants to compete with tools like ChatGPT and Google Gemini. But while it’s pushing innovation, it’s also creating serious privacy risks.
When you ask Meta AI a question, it replies just like any chatbot. But there’s a small “share” button that takes your chat and turns it into a post. You might think it’s saved privately. In many cases, it’s not.
Anyone can view these chats on the Discover feed. I found one where a teacher was discussing legal trouble with the AI. The bot responded with empathy. But the entire exchange was public. The teacher likely had no idea the world could see it.
This isn’t just one example. Other chats include questions about tax evasion, bowel problems, sexual health, and even emotional breakdowns. The Meta AI privacy leak makes these conversations visible—often with usernames attached. Sometimes, even audio recordings and images get posted.
What’s worse, many people use Meta AI through Instagram or Facebook. That means their real names might be tied to these private chats. And if your profile is public, so are your posts.
The chatbot replies kindly. It says your data helps improve its model. It promises to take steps to protect your privacy. But there’s no clear warning before you hit share. Most users don’t know the risks.
So what can you do? If you’re using the Meta AI app:
- Tap your profile picture
- Go to App Settings > Data & Privacy
- Under “Manage your information,” make sure your prompts are set to private
Also, don’t tap the share button unless you’re 100% sure. And if you use Meta AI on WhatsApp or Instagram, visit the Privacy Center. You can request that your messages not be used for training.
You can also delete chats. Just type /reset-ai in your conversation with the bot.
This Meta AI privacy leak shows how easy it is to expose personal information. Don’t let a casual chat become a public post.
Cybersecurity risks should never spread beyond a headline. Protect your social media accounts by using staying aware.
