Where the security score comes from
Eight channels, each asking one question of the published record and reading a named set of fields to answer it. They multiply rather than add — the factor beside a channel is what is left after the pressure it found, and open a channel to see every field it read, as it is stored. How the model works
Security fields
67 55 of them feed a channel
Established
29 37 left unknown, 1 not applicable
Completeness
41% the share the model could weigh
Citations
48 across 8 documents
Every value carries what it rests on, and the model weighs the four levels differently. Weight, then how many citations on this record sit at that level:
A weighs 1.00 · 11 hereB weighs 0.85 · 37 here
The weights are methodology.json’s own; a claim seen in the artefact itself counts for more than the same claim reported second-hand. The date at the end of a row is the day that value was last read.
A channel does not price every field it reads. The ones it did are marked along their edge and carry what they added; everything else was read and cost nothing. Only a channel that composes field by field can mark a field at all — the rest price themselves by the parameters printed above their rows.
Still open on this record.
Jurisdiction discrepancy. Browsec's own site states it is "a US-based company with its legal headquarters in Alexandria, Virginia," but the Virginia State Corporation Commission has no record of any entity named Browsec, and the USPTO trademark register lists the current owner of the BROWSEC mark as Browsec LLC of Novosibirsk, Russia. Which entity actually operates the service, and what if any US entity exists, is unresolved.
No independent audit of any kind. No no-logs audit, pentest, or infrastructure review is mentioned anywhere on the site, in its sitemap, or in general web search. The no-logs and "never tracks" claims rest entirely on the provider's own unverified word.
the rest of what is open — 126 more words
Retention specifics are unclear. The privacy policy names only an anonymous account ID, email, subscription data and a vaguely described "operational data" / "activity records of the device" kept indefinitely; it never states whether IP addresses or connection timestamps are logged at all, in either direction.
Server fleet is unclassified as physical or virtual. The location list names over 40 countries, including several under data-retention or interception mandates (Russia, Turkey, the company's own likely home jurisdiction), but nothing on the site says which of these are physical points of presence versus virtual locations.
Kill switch, split tunneling and DNS handling are not documented on any page found in this session (about_us, FAQ, platform pages), despite the FAQ covering other security claims (WebRTC/MITM protection) in similar detail.
On the record as a whole.
The about_us comparison table advertises a proprietary transport as "Custom VPN protocol: Xray/IPsec"; Xray-core is an open-source V2Ray-family obfuscation transport rather than a Browsec original, so it is coded as obfuscated (U07) rather than as a proprietary vendor protocol. The browser extension additionally offers "HTTP Proxy over TLS," which is a proxy rather than a full-tunnel VPN protocol and is not counted in U07.
The Terms of Service name Paddle.com as "Merchant of Record for all our orders except Russia," a carve-out consistent with the USPTO record placing the registered owner of the Browsec mark in Russia.
IntegrityHas it been caught saying one thing and doing another?16 fields read · 7 left unknown · 63.3 points of pressure×0.367
What this channel priced on this record:Lie 0.4Absolute claims 0.04Store declaration 0.15Owner transparency 0.25
Reported numbers plausible for this scalealso read by ignoranceP06_report_plausibility
Computed
×0.019 of a clean record completeness 41%
Where the drop comes from: Integrity. It is the channel this record's risk is concentrated in — on its own it takes a clean record to ×0.367, and its bar above is filled to that.
Exposure is recorded as medium and the integrity status as caught_lying.
Against the rest of the register, this record sits furthest above on Architecture — ×0.950 where the register averages ×0.900 — and furthest below on Jurisdiction, ×0.380 against ×0.680.
Every row above is one field of the published record, and every one of them can be challenged: every row carries Report, or raise something about the record as a whole. A correction with a source behind it is made the same day, and if it moves the score it moves the rank.
