Signal Alums Unveil Encrypted Spaces for Private Collaboration

Encrypted Spaces, reviewed by Matt Green and built by its team with a prototype application, is releasing code as a repository so researchers can help shape an architecture for end-to-end encrypted collaboration. The approach keeps a change log and uses zero-k
Matt Green has spent years thinking about encryption as more than a feature. So when he reviewed a white paper and a prototype application from the Encrypted Spaces project, his reaction was telling.
“They’ve built a system that’s kind of an extension of what end-to-end encryption can be. where you have an actual architecture for doing end-to-end encrypted collaboration. ” said Matt Green. a cryptography-focused professor of computer science at Johns Hopkins. “You can think of it as the Signal protocol for collaboration apps.”.
The promise is straightforward: private collaboration tools that work like modern group apps. without sacrificing the core principle that makes encrypted messaging feel trustworthy. But there’s a catch—Encrypted Spaces isn’t releasing a finished, ready-to-use app. Instead. it is putting out a code repository and inviting cryptography researchers and developers to review it. with the explicit goal of letting others eventually build encrypted collaborative apps.
“We want to make it so there’s no reason a developer wouldn’t want to make their application end-to-end encrypted, because it becomes so easy,” Trapp said.
That “so easy” ambition is where Encrypted Spaces starts to look less like a messaging upgrade and more like an attempt to rethink how encrypted software can coordinate large groups.
End-to-end encryption hits a wall when there are many people
Encrypted Spaces aims to address a key limitation of end-to-end encrypted apps: when the server can’t decrypt users’ data. any changes have to be handled on users’ devices. That model is relatively workable for a simple setup—an app that acts like a pipe between two phones. where each device holds keys to decrypt a conversation.
But in a collaborative platform, where dozens or hundreds of people work together, that “users’ devices do everything” approach becomes constraining. On a conventional unencrypted service, like Slack or Google Docs-style systems, the server can store information and manipulate it centrally.
Encrypted Spaces proposes a different model. An app built with it manages data from a centralized server and lets users collectively make changes to encrypted information while keeping it encrypted.
The mechanism is built around a change log: a record of every change to encrypted data that users make over time. That log can be shared with the app on every user’s phone or computer. Then the app can implement changes locally, keeping everyone’s version of the information synced and up to date.
How the server helps without ever seeing the content
The trick is the use of zero-knowledge proofs. In Encrypted Spaces, the server uses zero-knowledge proofs to prove to every user’s device that no changes are missing and no rogue changes have been made—without ever accessing the unencrypted data or the changes themselves.
The project also leans on a roll-up property of zero-knowledge proofs. Instead of forcing every device to process the entire history of the change log. the server can “roll up” the changes into a succinct proof that the current state reflects the entire history. Perrin described it this way: “The server can roll up the changes into a succinct proof that this current state reflects the entire history.”.
Perrin added: “It can convince you it’s applied the change log correctly without actually having to send it.”
Beyond syncing, the server also uses zero-knowledge proofs to oversee how devices manage cryptographic keys. Those keys are what allow only authorized users to decrypt and alter data.
That matters for group management. Encrypted Spaces includes capabilities for inviting new users and for provably revoking access if someone leaves the group.
It also gives users a choice about what history to share. People can decide whether to share the full history of the app, or whether a new invitee should be limited to new messages or data added after they entered.
A code release meant to be tested, not trusted blindly
Green’s review places Encrypted Spaces in a familiar lineage—he frames it as a cousin of the Signal protocol, but for collaborative apps rather than messaging and calls. The project itself supports that comparison by releasing code for scrutiny.
For now, the Encrypted Spaces group is not offering a single, ready-for-use application. Instead, it is inviting cryptography researchers and developers to examine the repository, with the idea that others can build encrypted collaboration apps later—without needing cryptography knowledge.
The architecture is designed to make “end-to-end encrypted collaboration” something developers can actually implement, not something reserved for specialists.
Where it lands now is an unusual stage for the kind of product people hope for when they hear “encrypted collaboration.” It’s a system with a prototype and a white paper. but the real work—review. iteration. and eventual building—sits in front of the research community and developers ready to take the repository apart and improve it.
Encrypted Spaces Signal protocol end-to-end encryption encrypted collaboration zero-knowledge proofs change log cryptography secure group apps
So it’s like Signal but for group stuff? Cool I guess.
I don’t get it, are they making it “encrypted spaces” so nobody can spy or what? The title sounds like a scam to me though, like paywall vibes. If it’s code, why not a finished app already?
“Signal protocol for collaboration apps” sounds like they’re just copying what Signal already did. Like ok but who’s hosting this and is it still basically a server holding the keys? Because if the server can’t decrypt then how does anything update, seems impossible unless it’s fake.
Honestly I’m more worried about the “change log” part than the encryption. Isn’t a change log still info even if it’s encrypted? Also “no reason a developer wouldn’t want to” like yeah sure, until it’s too hard and breaks everything. Sounds promising but I’ll wait until someone less academic releases an actual app.