Most public discussion of misinformation focuses on social media platforms. Twitter, Facebook, YouTube, TikTok: these are where researchers look, where platforms make policy, and where journalists report. But there is a large and understudied information ecosystem that operates almost entirely out of public view: private group chats. WhatsApp groups, Telegram channels, Signal conversations, Facebook Messenger, iMessage threads, and their equivalents on other platforms carry an enormous volume of information daily, and almost none of it is visible to the researchers, fact-checkers, and platform trust teams trying to track misinformation.
The dynamics of private group chats differ from public social media in ways that make misinformation harder to counter. Trust is higher in private groups because the participants are known to each other: family members, friends, colleagues, community members. A forwarded message from a cousin you have known your whole life carries implicit credibility that an anonymous social media post does not. The social cost of questioning or correcting information in a group chat is also higher. Pushing back on a claim in a family WhatsApp group can create interpersonal tension that most people would rather avoid. So misinformation that would get challenged in a public forum often circulates unchallenged in private ones.
The forwarding behavior that drives information spread in these environments amplifies the problem. WhatsApp allows users to forward messages to multiple groups simultaneously. During critical information events, elections, public health emergencies, natural disasters, messages can spread through interconnected group networks at enormous speed. WhatsApp responded to this by limiting forwarding to five groups at a time after research showed that bulk-forwarded messages were significantly more likely to contain misinformation. But the limitation applies only to the forwarding feature. Screenshots and copy-paste are unaffected.
The private nature of these environments creates a research blind spot. Platform researchers can study what happens on their public platforms because the data is accessible. Group chats are encrypted, ephemeral, and not indexed. Fact-checking organizations cannot see the claims circulating in private groups and cannot intervene even if they could. The people who could intervene are the group participants themselves, and most have neither the training nor the motivation to fact-check claims that come from people they trust.
FactSignal’s investigation of addressing misinformation in encrypted channels, without compromising user privacy, finds that the options are narrow. WhatsApp’s forwarding limits helped reduce the most viral spread. Media literacy campaigns can build individual resistance to false claims. Some organizations have set up tip lines where users can forward suspicious messages for fact-checking. But the fundamental tension remains: you cannot fact-check what you cannot see, and encrypted group chats are designed to not be seen.
For individuals, the practical step is developing the habit of not forwarding things you have not verified, even, especially, when they come from people you trust. The trustworthiness of the sender is not the same as the trustworthiness of the information. A well-meaning relative forwarding an inaccurate health claim is doing more damage than a bot on Twitter, because the claim arrives wrapped in personal credibility. Unwrapping it takes effort. But leaving it wrapped takes a toll on the collective accuracy of the group’s shared understanding.
