What Mullvad's Warrant Canary Actually Promises — and the Legal Theory That Has Never Once Been Tested in Court

The number of times a warrant canary has been upheld as legally effective in a United States court is zero. Not "rarely." Zero. The whole mechanism rests on a theory of compelled speech that no judge has ever ruled on. So before you treat a canary as protection, you need the vocabulary to know what it is — and what it is not. You do not need to be a lawyer or a cryptographer here. You need these nine terms, in order, each building on the last.

Warrant Canary

A warrant canary is a public statement that an organization has *not* received a particular kind of secret legal demand — published precisely so that its removal signals the opposite.

Why it matters in practice: the logic is inverted, and that inversion is the entire point. A provider cannot legally tell you "we got a secret order." But it can publish, every month, "we have received no secret orders." When that sentence stops appearing, you are meant to read the silence. For a VPN provider whose adversary is a subpoena, this is the only signal the gag order theoretically permits.

Concrete example: Mullvad publishes a canary statement asserting it has received no national security letters, gag orders, or warrants it cannot disclose. If you are the salaried sysadmin routing your company's remote traffic through it, the canary is not the firewall — it is the smoke detector. It tells you when to leave, not whether the building is safe.

Gag Order

A gag order is a legally binding prohibition that bars the recipient of a legal demand from disclosing that the demand exists.

Why it matters: the canary only has meaning *because* gag orders are real and enforceable. A National Security Letter in the US can carry an indefinite nondisclosure obligation under 18 U.S.C. § 2709. The provider is muzzled — it physically cannot speak the truth without committing a crime. The canary is an attempt to route around that muzzle by speaking the *absence* of the thing, in advance, while still legally free to do so.

Concrete example: if your threat model includes a passive ISP collecting browsing metadata, a gag order is irrelevant to you — your adversary isn't subpoenaing your VPN. But if your threat model is a state actor with MLAT authority, the gag order is the precise mechanism that makes your provider go quiet. The canary is built for *that* reader, not the first one.

National Security Letter

A National Security Letter (NSL) is an administrative subpoena issued by the FBI without prior judicial approval, typically demanding subscriber or transactional records.

Why it matters: the NSL is the archetypal threat a canary is designed to flag, and its defining feature is the attached nondisclosure requirement. No judge signs off before it lands. That absence of judicial review is exactly why the canary exists as a workaround — there is no public docket, no court record, nothing the user could otherwise see.

Concrete example: the famous real-world canary event was Apple's. Its 2013 transparency report contained language stating it had never received a Section 215 order; the language quietly disappeared from the 2014 report. Apple never confirmed why. That is a canary functioning exactly as designed — the removal *is* the message, and the company stays silent on the rest. A VPN canary aims to reproduce that exact mechanic for a much smaller, much more targeted set of demands.

Compelled Speech

Compelled speech is the legal doctrine that the government generally cannot force a private party to make a *false* affirmative statement.

Here is the concession, and it's a real one: the strongest argument for canaries lives here. The First Amendment theory is that while a gag order can stop you from speaking, forcing you to *keep publishing a lie* — "we received no orders" when you did — is a different and more constitutionally fraught act. That distinction is genuinely sound on paper. Now the teardown: no court has ever ruled on it in the canary context. The doctrine that protects the canary is untested precisely where it matters, and a sufficiently motivated prosecutor could argue that removing the statement is itself a prohibited disclosure. You are trusting a theory, not a precedent.

Concrete example: this is why a careful provider words the canary as a present-tense factual assertion it can simply *stop renewing* — not as an active announcement of an order, which would more clearly violate a gag.

Dead Man's Switch

A dead man's switch is an automated mechanism that triggers an action — or stops a recurring one — when a human fails to actively prevent it.

Why it matters: a canary that depends on someone manually deleting a sentence is fragile. A dead man's switch flips the failure mode. The default becomes *no statement*, and a living, un-gagged human must actively re-sign and re-publish to keep the canary alive. If they're arrested, compromised, or legally frozen, the canary lapses on its own. Silence becomes the automatic state, not a deliberate act someone might be ordered not to take.

Concrete example: an operational canary tied to a cadence works as a soft dead man's switch. If the page is supposed to refresh by the first of every month and the calendar rolls past with no fresh signature, the switch has effectively fired — nobody affirmatively renewed it. That is the design intent: make *inaction* the alarm, because inaction is the one thing a gag order cannot compel.

Cryptographic Signing

Cryptographic signing attaches a digital signature to the canary, proving it was produced by the holder of a specific private key and not altered since.

Why it matters: an unsigned canary is worthless. Anyone who compromises the web server can edit a plain HTML sentence and you would never know. A PGP or similar signature — verified against a key you obtained independently — proves authorship and integrity. Per the public-key model the signature also resists the scenario where an attacker keeps publishing a forged "all clear" after seizing the infrastructure but not the offline signing key.

Concrete example: a properly signed canary includes the statement *and* a detached signature you check against a long-published fingerprint. If you are the IT professional onboarding a VPN for a regulated workload, your verification step is not "did I read the page." It is "did the signature validate against the key I cached six months ago." A signature that suddenly verifies against a *new* key is itself a red flag worth escalating.

Update Cadence

Update cadence is the fixed, pre-announced interval at which the canary must be re-signed and re-published to remain valid.

Why it matters: cadence converts the canary from a static claim into a time-bound heartbeat. Without it, you cannot distinguish "nothing has happened" from "the page is just old." A stated cadence — monthly, quarterly — gives the silence a deadline. Miss the deadline, and the absence becomes evidence rather than ambiguity. The shorter the cadence, the tighter the resolution on *when* something changed.

Concrete example: say the canary commits to a fresh signed statement within the first seven days of each quarter, and references a recent, unpredictable public datum — a block-height, a headline — to prove it was signed *after* a given moment and not pre-generated under coercion. If Q3 arrives and the page still shows the Q2 datum, you do not wait for an announcement. The cadence already told you what to do.

Jurisdictional Reach

Jurisdictional reach is the set of legal frameworks under which a provider can actually be compelled — which depends on where it incorporates, where its servers sit, and what treaties bind both.

Why it matters: a canary is only as meaningful as the legal pressure it can detect, and that pressure is jurisdiction-shaped. Here is the bridge that most reviews skip. What the United States accomplishes through an NSL and 18 U.S.C. § 2709's gag provision, a European state often accomplishes through a mutual legal assistance treaty request routed under the Budapest Convention on Cybercrime (2001) — different instrument, same effect on your provider. A Sweden-based operator is outside direct NSL reach but squarely inside EU mutual-assistance channels and Schengen information sharing. "Not in the US" does not mean "not reachable."

Concrete example: a provider can run a flawless canary and still be served through a treaty mechanism its canary language never anticipated. Read what the canary actually disclaims — does it cover foreign demands, or only domestic ones?

Canary Verification

Canary verification is the user-side process of independently confirming that a canary is current, correctly signed, and unchanged in its core assertions.

Why it matters: a canary you never check is theater. Verification is where the mechanism either earns its keep or fails silently — and the failure is silent by design, which means the burden sits with you. This is the operational core of the whole subject, and it is the part marketing pages bury. The provider builds the signal; you are the receiver, and an unmonitored receiver detects nothing.

Concrete example: practical verification is a recurring task, not a one-time read. You cache the signing key's fingerprint independently. You note the committed cadence. Then, each interval, you confirm three things — that a fresh statement exists, that its signature validates against the *same* key, and that the disclaiming language did not quietly narrow. A canary whose wording shrinks from "no orders of any kind" to "no orders we are aware of" has changed its meaning without missing a single publication date. That edit, not the deletion, is the one most readers never catch.

FAQ

Does Mullvad actually run a warrant canary, and what does it cover?

Mullvad publishes a signed statement asserting it has not received warrants, gag orders, or national security letters that would compromise users. Treat the published scope as the literal boundary of the promise. Read whether it disclaims only certain demand types or all of them, and confirm it is cryptographically signed and current against its stated cadence before relying on it. The statement's wording is the contract, not the marketing copy around it.

Is a warrant canary legally enforceable?

No court has ever ruled a warrant canary legally effective, in the US or elsewhere. The mechanism rests on a compelled-speech argument — that the government cannot force a provider to keep publishing a falsehood — which is plausible on First Amendment grounds but entirely untested in the canary context. You are trusting an unproven legal theory. A motivated prosecutor could argue that removing the statement is itself the prohibited disclosure a gag order forbids.

What's the difference between a warrant canary and a no-logs policy?

They answer different questions. A no-logs policy claims the provider does not retain data that could be handed over; a warrant canary signals whether a secret demand has arrived. A no-logs claim is only meaningful when independently audited for scope and timing — "audited" without a stated scope means little. The canary tells you about legal pressure; the logging policy tells you what that pressure could actually extract. You want both, verified separately.

How do I verify a warrant canary myself?

Independently cache the signing key's fingerprint from a source other than the canary page. Note the provider's committed update cadence. Then, each interval, confirm three things: a fresh statement exists, its signature validates against the same cached key, and the core disclaiming language has not quietly narrowed. A signature that suddenly validates against a new key, or wording that softens, matters as much as an outright deletion.

If the canary disappears, does that mean my data is compromised?

Not necessarily — and that ambiguity is the mechanism's central weakness. A lapse can mean a legal demand, but it can also mean operational neglect, a key-rotation mishap, or an administrative error. The canary signals *investigate*, not *confirmed breach*. It is a smoke detector, not proof of fire. The correct response is to treat the silence as a prompt to reduce your exposure, not as a verdict about what already happened to your traffic.

Does a warrant canary protect against foreign governments?

Only to the extent its language and the provider's jurisdiction account for them. A canary worded around domestic US demands may say nothing about a request routed through a mutual legal assistance treaty under the Budapest Convention (2001). A provider outside US reach can still be compelled through EU mutual-assistance channels. Check whether the canary explicitly disclaims foreign and treaty-based demands, not just domestic warrants — the gap is where most assumptions break.

How often should a warrant canary be updated?

Frequently enough that silence has a sharp deadline, and tied to an unpredictable public datum to prove recency. Monthly or quarterly cadences are common; shorter intervals give finer resolution on *when* something changed. The canary should reference a recent, unforeseeable public value — a news event or block-height — so it cannot have been pre-signed under coercion. Without a stated cadence, you cannot tell "nothing happened" from "nobody bothered to update the page."

Should the existence of a warrant canary change which VPN I choose?

It should inform, not decide. A canary is one signal among several — alongside independently audited no-logs claims, jurisdictional exposure, and protocol architecture. Providers such as ProtonVPN, NordVPN, ExpressVPN, and Surfshark each present different combinations of audit history and jurisdiction, and a canary only matters if your threat model actually includes secret legal demands. If your adversary is a passive ISP, the canary is irrelevant to you. Match the tool to the adversary first.