Is Flotilla Private?

Part of our guide to running an online community you actually own.

Digital privacy matters a great deal, and it’s also complicated. Calling something “private” can mean different things to different people — anything from “trust me bro” to full end-to-end encrypted communication.

Flotilla covers two main areas where the requirements and guarantees differ: community spaces and direct messages.

Space Privacy

Flotilla takes a pragmatic approach. Rather than leaning on encryption, it protects spaces mainly through self-hosted community infrastructure. This gives organizers more control over exactly what stays private and who has access, instead of setting the rules for your community in stone with encryption.

So a space is as private as the policies of the relays that host it. If an administrator wants the space to be invite-only and private, they can enforce that. If they want community content to be read-only for non-members, that’s an option too. The important thing is to make the expectation clear to members.

A group is only as private as its least trusted member. Every time someone new joins, the guarantee that messages won’t be shared out of context gets weaker. Encrypting a 50-member group would be overkill, given how weak its privacy is likely to be along that social dimension.

It’s also worth mentioning that Flotilla is built on the Nostr protocol, which uses digital signatures to verify messages. That gives it weak deniability: if content leaks out of a group, it can be proven authentic, because the author signed it.

For this reason we describe Flotilla’s groups — particularly those hosted by Coracle Hosting or using the Zooid relay implementation — as “non-public” rather than “private”. Flotilla isn’t a solution for journalists or activists; it’s for small communities.

What Is Public About You

Before we split things into spaces versus direct messages, it’s worth being clear about what’s public about you on any space you join, no matter how tightly its policies are set.

Your public key — your npub — is exactly what the name says: public. It isn’t a secret, and it was never meant to be one. Anyone can see it, and it’s how other people and other apps recognize you as the same person across Nostr. Every public message you send, along with your profile (display name, bio, avatar, and the rest), is signed by that key and readable by anyone who can read the relay hosting it. Your membership in a space and the reactions you leave on other people’s posts work the same way: they’re visible to anyone who can read that relay, not hidden from other readers just because they weren’t posted in the main chat. This doesn’t mean every relay is open to outsiders. As covered above, relays are closed by default, and non-members only see what an operator chooses to expose. But whatever an operator does expose, your own activity within it gets no extra protection.

It’s also worth treating anything you publish as effectively permanent, even where a delete option exists. Flotilla can delete a message you posted, and a self-hosted relay running zooid can be set to expire events automatically or let an admin hard-delete something — but none of that is a guarantee. Deletion on Nostr is a request, not a command. Neither Flotilla nor a relay operator can force every other relay, or anyone who already copied your content, to honor it. See Can I Delete My Account or Data? for exactly what Flotilla’s delete tools do and don’t accomplish.

One more thing shapes what you reveal: some relays ask Flotilla to prove who you are before serving certain content. Flotilla does this automatically only for relays already in your own relay lists. Anywhere else, a setting decides whether it authenticates on your behalf. Leaving it off lets you read from an unfamiliar relay without the operator learning your identity; turning it on trades that away for relays that require it. See Privacy Settings: Limiting What You Leak for that toggle and the rest of Flotilla’s privacy controls.

Direct Messages

Flotilla takes a different approach for direct messages, which are a much more private form of communication. Because a DM rarely involves more than a handful of people (Flotilla supports small group chats as well as one-on-one conversations), it makes sense to protect them with encryption — especially since DMs are often stored on public infrastructure.

So Flotilla uses the NIP 17 standard for end-to-end encrypted direct messages. As a result, service providers can’t see your conversations, and you also get deniability and metadata protection.

What Flotilla DMs don’t have is forward secrecy or post-compromise security. In plain terms, if your Nostr identity is ever compromised, an attacker could gain access to old messages.

This is a trade-off between user experience and security: those stronger guarantees would also mean your messages weren’t available when you signed in on a new device.

The Future of Messaging

In the future, Flotilla may support an emerging standard called Marmot, pioneered by the team at WhiteNoise. This new protocol has all the strong encryption features mentioned above, plus the ability to host larger groups.

The standard is very new, and porting Flotilla’s more sophisticated group-management features to it would take a long time, so for the moment we aren’t actively adopting it, though we hope to in the future. In the meantime, if you’re looking for stronger privacy guarantees, try WhiteNoise.