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
40 26 left unknown, 1 not applicable
Completeness
57% the share the model could weigh
Citations
56 across 12 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 · 5 hereB weighs 0.85 · 41 hereC weighs 0.70 · 10 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.
The yearly price, corrected. The pricing page's Essential-tier yearly card advertises "$179.88" struck through against an actual first-year charge of $46.68 (equivalent to $3.89/mo), with a separate line stating it "Renews Yearly at $99.99"; the previous record had stored that renewal figure in U04b_price_yearly rather than the first-term charge. U04_price_monthly=14.99 is unaffected: the Essential month-to-month card carries no promotional discount and stood on re-read.
Founding year and user scale. M07 (launch year) and M08 (user scale) are unestablished: no page read this pass states a founding year or a subscriber count, and the /about/ URL guessed for this run 404s rather than redirecting to a real page.
the rest of what is open — 162 more words
Audits behind a login. V04 (fleet coverage), V07 (assurance standard) and V15 (assurance strength) stay unknown because both named audits (Schellman 2025, Leviathan Security 2022) are gated behind the customer account portal per AUDITOR.md's login-gated-report rule — the trust page's own list of what IPVanish does not log is the provider's marketing claim, not the audit's finding, so V11 is recorded not_stated rather than borrowing that list.
Looked for, and not found. C08 (server hardware ownership), C12 (device identifier) and U02 (the server-country list) were not found on any page opened. P04 (mandate-response) and P07 (warrant canary) were looked for and not found on the privacy policy or trust pages. X10 (affiliate payout tier) was not checked.
The legal entity that is never named. The exact registered legal-entity name (an LLC/Inc. suffix) is never stated in the privacy policy — only "IPVanish"/"the Company" and a Ziff Davis byline — so M02 records that absence directly rather than filling in a plausible-sounding name.
On the record as a whole.
The Essential-tier yearly plan (U04b_price_yearly=46.68) renews at $99.99/year after the first term, and the Essential month-to-month plan (U04_price_monthly=14.99) autorenews at the same rate; both read from the live pricing page on 2026-09-13.
IPVanish is this catalogue's second and cleaner instance of X02=denied_then_proven, the first found inside a sworn federal document rather than reconstructed from marketing copy. In United States v. Vincent Gevirtz (S.D. Ind., No. 1:16-mj-00487-TAB, Criminal Complaint and affidavit of SA Michael Johnson, filed 2016), the government quotes Highwinds Network Group's (IPVanish's then-operator) own written response to a first summons, dated 2016-05-26: "we do not log any usage information... we do not have any information regarding the referenced IP" (affidavit para. referencing the first summons response).
the rest of what is open — 432 more words
One month later, responding to a second, more detailed summons dated 2016-06-21, Highwinds produced a full connection-log sequence (VPN-connect / IRC-connect / VPN-disconnect timestamps), the customer's real name, email and account/subscription status, and — per the affidavit's next paragraph — the actual originating Comcast IP address tied to those connections.
The denial and the production are both quoted directly from the affidavit's own text (confidence A on both P02/P03 and X01/X02), not reconstructed from a provider blog post the way PureVPN's record had to be. R11_evidence_voiding is the reason this event does not sit at full undecayed weight on the current owner: IPVanish's operator in 2016 was Highwinds Network Group / Mudhook Media, not the current owner.
Ownership passed to StackPath in 2017, then to J2 Global (renamed Ziff Davis, Inc. in 2021) in a deal TechRadar (published 2019-05-07, read directly) dates via an editor's note that PCMag itself — a Ziff Davis-owned title — published disclosing the conflict of interest: "IPVanish and StrongVPN are owned by J2 Global".
That same disclosure is recorded in X11_review_media_disclosure=disclosed: it is PCMag's own note, described by an independent outlet, not IPVanish's self-report. M16_control_since is set to 2019-05 (month precision only — TechRadar's publication date, not a located acquisition-closing announcement) specifically so R11 discounts the 2016 event toward the retention floor rather than holding it at full price against an owner who did not exist at the time of the incident.
Ziff Davis, Inc. is a publicly traded NASDAQ company (ticker ZD, confirmed live on ziffdavis.com) whose Technology segment describes itself as the "CNET Group portfolio" and whose Cybersecurity segment names VIPRE Security Group; PCMag itself could not be independently read this session (pcmag.com returned HTTP 403 to curl and is blocked outright at the WebFetch tool level for this host), so the PCMag-Ziff Davis ownership tie rests on TechRadar's description rather than a PCMag masthead page, and M10_owner_other_business's review_directory entry carries that same C-confidence limitation.
IntegrityHas it been caught saying one thing and doing another?16 fields read · 5 left unknown · 86.5 points of pressure×0.135
What this channel priced on this record:Lie 0.85Bought coverage 0.1
Reported numbers plausible for this scale
Computed
×0.010 of a clean record completeness 57%
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 ×0.135, and its bar above is filled to that.
Exposure is recorded as low and the integrity status as caught_lying.
Against the rest of the register, this record sits furthest above on Silence — ×1.000 where the register averages ×0.927 — and furthest below on Integrity, ×0.135 against ×0.658.
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.
