Hands-on with Pretty Good Phone Privacy, a data service for Android that aims to give users an additional layer of privacy when using mobile phone networks
Joseph Cox / VICE : Tweets: @josephfcox , @josephfcox , @lorenzofb , and @josephfcox Tweets: Joseph Cox / @josephfcox : On the backend, PGPP gets services from Telna, a Canadian company, which in turn gets the IMSIs from Play, a Polish telecom (when I tried to sign up to Signal, it anticipated a Polish number). I used Wickr as that doesn't require a phone number for signup https://www.vice.com/... https://twitter.com/... Joseph Cox / @josephfcox : But really the biggest questions on whether people will want this are: where else can third-parties try to grab data on the users (just go to underlying telecos?) and how would PGPP respond to data requests? PGPP says it has little info to provide to LE https://www.vice.com/... https://twitter.com/... Lorenzo Franceschi-Bicchierai / @lorenzofb : Someone made a phone specifically for Joe. https://twitter.com/... Joseph Cox / @josephfcox : New: I tried Pretty Good Phone Privacy (PGPP), the pseudo-phone network that aims to delink your ID from your phone usage. Cycles through new IMSIs on demand, routes all device internet traffic through two hops, data-only eSims. It's interesting for sure https://www.vice.com/... https://twitter.com/...
Context & Ripple Effects
Joseph Cox's hands-on puts working details on Invisv's Pretty Good Phone Privacy beta that earlier coverage of the Android service left abstract: on the backend, PGPP buys connectivity from Telna, a Canadian company, which in turn sources IMSIs from Play, a Polish telecom — so a user's traffic is separated from their identity, but their number can still betray a Polish origin, as when a Signal signup anticipated one.
The piece sits in a longer arc VICE has tracked: from criminal-grade anonymized SIMs like the so-called white SIMs and the Phantom Secure network dismantled by the FBI, toward consumer-facing privacy carriers — a category Phreeli later pushed further with ZIP-code-only signup and zero-knowledge proofs.
First-order effects
- Android users on PGPP get a working layer that decouples them from network-level identifiers today, but the hands-on exposes the practical friction: services that demand phone verification, like Signal in Cox's test, surface the Polish numbering behind the service.
- Apps that don't require a phone number, such as Wickr, become the de facto pairing for PGPP users — the service's real-world utility depends on which messengers tolerate its non-native numbers.
Second-order effects
- The Telna-to-Play supply chain shows privacy carriers are resellers of wholesale connectivity, so their privacy claims are only as strong as every upstream hop — pushing rivals like Phreeli to compete on minimizing what they collect at signup rather than just encrypting in transit.
- Verification-dependent platforms face a choice as anonymized-number users arrive: tighten checks and alienate privacy buyers, or accept foreign-origin numbers and absorb the fraud risk.
Third-order effects
- If the model holds, mobile access stratifies into a commodity wholesale layer (Telna-style aggregators sourcing identifiers from carriers like Play) beneath branded privacy frontends — shifting power over user identity to whoever controls the identifier-mapping middle.
- The Phantom Secure precedent looms over the category: consumer privacy carriers that genuinely sever identity from usage will eventually draw the same law-enforcement scrutiny that ended its criminal-market ancestor.
The trend: Consumer mobile privacy is becoming a product category built on resold wholesale carrier infrastructure, where the depth of anonymity is set by the backend supply chain rather than the marketing.