Olvid¶
- Best suited for
- Avoid sharing a number with the contact; verify cryptographic identity. [1] [2] [3] [4]
- Main strength
- No mandatory number; documented TURN-only calls. [1] [2]
- Main weakness
- No network anonymity; calls and multi-device require a licence. [1] [2]
- Account identifier
- Cryptographic identity; no mandatory phone or email. [1] [2]
- Android
- Yes, client listed in the study. [1]
- Confidence and scope
- Scope-limited documentary evidence; production metadata and distributed binaries not experimentally verified. [1]
What your contact learns¶
- What the contact learns
- The contact receives messages and the profile you share, and can keep a copy. [1] [2]
What the provider learns¶
- What the provider learns
- Routing and network information; actual production retention and access unknown. [1] [2]
Other observers¶
- What the network / ISP learns
- The network sees destinations, timing and volume; Tor or a relay changes the observer. [1] [2]
- Apple / Google exposure
- Store, installation and push can create metadata; exact profile linkage depends on the OS and was not tested here. [1] [2]
- Public law-enforcement evidence
- No individual disclosure case established here; this does not prove an absence of requests or data. [1] [2]
Encryption and keys¶
- E2EE
- Formally studied protocol under limited assumptions; not an audit of the entire 2026 system. [1] [2]
- Forward secrecy
- Unknown: not established for the version and configuration in question.
- Post-compromise security
- Unknown: not established for the version and configuration in question.
- Key verification
- Verify identity or fingerprint in person / through a separate channel. [1] [2]
Metadata and identity¶
- Metadata exposure
- Routing keys, sizes and timing visible in the formal model. [1] [2]
- Global user ID
- Cryptographic identity; server sees routing keys in the analysed model. [1] [2]
- Username / discovery
- Unknown: not established for the version and configuration in question.
- Push notifications
- Unknown: not established for the version and configuration in question.
Calls, IP and groups¶
- IP exposure to peer
- Documented TURN-only calls: IP hidden from the peer, visible to the relay. [1] [2]
- Relay / TURN
- Documented call candidates are TURN-only; no current packet capture performed. [1] [2]
- Groups
- Unknown: not established for the version and configuration in question.
Devices and backups¶
- Multi-device
- Licensed multi-device; old history not automatically transferred. [1] [2]
- Backups
- Unknown: not established for the version and configuration in question.
Audits, limitations and public cases¶
- Audits
- Historical CSPN/Synacktiv work and 2026 formal analysis; distinct scopes and versions. [1] [2] [3]
- Source code and binaries
- Repositories and component-specific scopes recorded in Phase 2; no current binary rebuilt bit-for-bit here. [1] [2]
- Main weakness
- No network anonymity; calls and multi-device require a licence. [1] [2]
- Public law-enforcement evidence
- No individual disclosure case established here; this does not prove an absence of requests or data. [1] [2]
Configuration and unknowns¶
- Recommended configuration
- Verify identity / fingerprint out of band; assess licensing and history before adding a desktop. [1] [2] [3] [4]
- Remaining unknowns
- Current network behaviour, distributed binaries and forensic evidence remain incomplete; check version, OS and configuration. [1] [2]
Understand the terms¶
End-to-end encryption — E2EE
Participants’ devices control content keys. The transport server should not read ordinary messages. Learn more
Metadata
Data around the exchange: accounts, IP, timing, volume, devices or groups. Learn more
Pseudonymity
Use an identity different from your civil identity. It can still be linked to you. Learn more
Network anonymity
Reduce the ability to link traffic to your network address or its destination. Learn more
The detailed evidence is maintained in the French Phase 2 register. The translation is a restatement, not new evidence.