Briar

Briar is an open-source, censorship-resistant messaging app that works without any central server. Messages are synchronized directly between users' devices and are end-to-end encrypted, so there is no provider to subpoena, block, or breach. Built for activists, journalists, and anyone facing surveillance or network shutdowns, it is one of the few messengers designed to keep functioning when the internet itself is cut.

How it connects

When the internet is available, Briar routes traffic through the Tor network, hiding not just message content but also metadata about who is talking to whom — an observer of your connection sees Tor traffic, not a messaging pattern. When the internet is down, it falls back to direct Bluetooth or Wi-Fi links between nearby devices, and it can even move messages by removable storage such as a USB stick — literal sneakernet, encrypted end to end. This multi-transport design means a Briar conversation can survive an internet blackout that would silence conventional apps entirely. The peer-to-peer architecture does impose one honest cost: with no server to hold messages, both parties must be reachable for delivery over Tor. The project's answer is the Briar Mailbox, a companion app you run on a spare Android device that stays online and stores incoming messages for you, encrypted, until your main device collects them — a buffer you host yourself rather than a cloud you must trust.

More than direct messages

Briar also supports private groups, public forums, and blogs, all synchronized over the same mesh of encrypted connections. Content propagates from contact to contact, so a forum post can spread through a community even if no single member has internet access — each phone carries the data onward at the next encounter. Contacts are added by exchanging links or scanning QR codes in person, which doubles as key verification: there is no phone-number registry, no discovery server, and no central directory of who exists.

Why it matters for sovereignty

Fitting Briar to a threat model honestly means naming what it does not do. It will not hide that your device is running Tor from a network observer, it cannot protect a conversation if the endpoint device itself is compromised, and its in-person contact exchange — a security strength — is a real logistical constraint for contacts you can never meet. Message history lives only on the devices involved, so losing a phone without a backup loses the archive. None of these are flaws so much as the honest costs of serverlessness, and they define its niche: Briar is what you reach for when the network is hostile or absent and the people you need to reach are known to you — a communications tool with the same self-reliant assumptions as an off-grid power system.

Briar removes the central point of control and observation that defines mainstream messengers — no cloud backup, no server-side contact graph, no company that can be compelled to hand over what it never had. That makes it a strong fit for the same threat model that drives self-custody of Bitcoin: minimize trusted third parties, hold your own data, assume infrastructure can fail or be turned against you. The trade-offs are the usual price of purity — battery drain from Tor and Bluetooth duty, Android-first availability, and no cross-device sync — so for most people Briar is the resilient fallback in the toolkit rather than the daily driver. Slot it into a layered plan: a mainstream encrypted app built on the Signal protocol for everyday end-to-end encryption, Briar for blackouts and high-risk contexts, radio tools like Meshtastic for short-range off-grid mesh and Winlink for long-haul radio email. For an alternative architecture that removes even random user identifiers, see SimpleX Chat — and for deciding which tool fits which risk, start with threat modeling.