Archived snapshot: Substack · Is It Training on Me?

← Substack · AI Addendum

This is an extract from an archived copy, fetched 10 September 2026. Substack didn't author this page for us. It's a snapshot we captured from https://support.substack.com/hc/en-us/articles/50891130623508-How-can-I-detect-AI-on-Substack. We hold the whole page and publish this part of it: the sentence a verdict rests on, with the text either side, so you can see it hasn't been lifted out of an exception.

SHA-256
55EE4CB9…CCB9

The passage, in context

Paragraph 3 of 38

How can I detect AI on Substack?

Scan for AI text estimates whether text was written by a human or AI-assisted. It works on posts and notes published on or after July 21, 2026, and uses Pangram to assign a percentage of how much of the text is human-written or AI-assisted.

Note: Neither Substack nor Pangram uses publisher content to train generative AI models. Substack offers a Block AI training toggle in your publication's Settings that signals to external AI crawlers not to use your content for training models.

The Scan for AI text tool can be used in the following surfaces and content:

The Substack Reader (Web): Click on a post from your Subscriptions to open the reader.

Substack iOS App (Android coming soon): Check for AI on notes and posts on the app.

Checking the rest of it

The complete page is kept here and isn't republished. It's Substack's copyrighted document, and an extract is what a reader needs to check a quotation. The SHA-256 above is of that whole snapshot: fetch the page yourself, hash it, and you can tell whether ours has been altered without having to trust us.

If you work for Substack, or you are researching this and need the full capture, ask us for it and we will send it. If you believe a quote here is wrong or out of date, send us the paragraph. The correction gets published beside the clause.