232 APPS TRACKED · 227 CLAUSES ON FILE · 38 WITH NO CLAUSE TO QUOTE
RECORD OPENED 5 OCT 2026 · 1 CLAUSES ON FILE
comms

Does Signal train on your data?

Signal's own policy says it cannot decrypt or access the content of your messages or calls. Not a promise not to train on them — a statement that they are not readable to Signal at all.

LAST CHECKED
5 Oct 2026 · 5 days ago
CONFIDENCE
high
RIGHT OF REPLY
Not yet asked
Verdict
A
Cannot train on you
Your content is not readable to them

Evidence 1 — All accounts
Privacy Policy · All accounts
EVIDENCE 01 OF 01
"Signal cannot decrypt or otherwise access the content of your messages or calls."
CHECKED
1 Aug 2026
STILL LIVE AS OF
5 Oct 2026
APPLIES TO
All accounts
JURISDICTION
global
HOW IT WAS TAKEN
Internet Archive copy
SHA-256
F043C519…54F1

How we read it: Signal's policy contains no training, machine-learning or AI language anywhere: searched on 1 August 2026 for train, training, fine-tune, machine learning, artificial intelligence, AI model and large language model, with no match in a training sense. That absence is why this read UNCLEAR until the A rule was written, and it is the wrong question to ask of a service that cannot read the content: the same document says three separate times that Signal cannot reach it. What Signal does hold is stated plainly there and sits outside this grade's scope — the phone number you register, and "randomly generated authentication tokens, keys, push tokens" held on its servers to route messages.

What each plan can refuse

The same questions, asked of every plan we could identify. Every cell comes from a quoted clause or a documented search - see the evidence above.

All accountsTRAINS: CANNOT
OPT-OUT
Nothing to opt out of
RETENTION
Not stated
EXPOSURE
5 / 100
DE-IDENTIFIED
Not stated
HUMAN REVIEW
No

How to opt out of Signal AI training

Plan by plan, from what Signal's own policy says about refusing. Where it publishes the steps, they're listed here with the limits they carry. Where there's nothing to switch off, or nothing graded, this says so instead of inventing a menu path.

All accounts

There is nothing to opt out of. Signal's own policy says it can't reach the content, so training on it isn't something a setting controls.

Each step's source and archived copy →

What this policy has done since we started watching

1 CAPTURE SINCE 1 AUG 2026

The record starts here. Signal's policy was first captured on 1 Aug 2026, and there's no earlier copy to compare it against, so this verdict has no history behind it yet. From now on every re-read is diffed against that copy, and anything that moves appears here.

Everything we read

DOCUMENT
CHECKED
SNAPSHOT
Privacy Policyquoted
1 Aug 2026

Vendor response

From the company? What a reply does and what moves a grade.

We haven't yet written to Signal about this verdict, and they haven't disputed, clarified or announced a change to any clause quoted above. When either happens it's published here, word for word.

Report a wrong clause

Compared with Telegram and Threema · Writing about Signal? Embed this verdict as a card that stays current.

Show this verdict on your site

Signal is graded A, and a company that's earned that can say so with the verdict itself. The image below links back to this page and shows the live grade. If the grade changes, the badge changes the same day, wherever it's embedded. It's free and it isn't for sale.

Signal is graded A on Is It Training on Me