Substack Adds Pangram-Powered AI Detection to Fight "Claudefishing"

In July 2026, Substack launched a built-in AI text detection feature powered by the Pangram service. Readers now get a button in the interface that flags posts allegedly written by a neural network. The platform announced the feature on its own blog in a post titled "Against Claudefishing."

What Is Claudefishing

Claudefishing is a term coined by Substack for the moment a reader realizes that a post they admired for its "living human voice" was actually generated by a bot. The word combines the name Claude — Anthropic's chatbot — with "catfishing," the internet deception of posing under a fake identity.

Substack put the term right in the headline of the manifesto that accompanied the feature's launch. The platform positions itself against venues where an author's voice is increasingly imitated by a machine, and makes the "humanness" of text an explicit competitive advantage.

• Term: Claudefishing — from "Claude" + "catfishing"
• Detection engine: third-party service Pangram
• Announcement format: Substack blog post "Against Claudefishing"
• Mechanism: an indicator button next to the post; the decision is up to the reader

How the Detection Works

The check runs on Pangram, a third-party service that specializes in recognizing AI-generated text. Substack did not build its own detector; it integrated someone else's. An indicator appears next to a publication, and readers can use it to judge whether a human or a model wrote the text.

Substack leaves the decision to the audience. The platform does not automatically block or delete suspicious posts; it puts a verification tool in readers' hands. This is a deliberately soft approach — a label, not a ban.

Why It Matters for Authors

Substack has for the first time officially acknowledged the problem of AI content inside its ecosystem — a service built precisely on trust in a personal authorial voice. For a newsletter platform where subscribers pay for a specific author, a "bot disguised as a human" strikes at the core asset: the reader's belief that a real person is on the other side of the screen.

"Claudefishing is the moment you realize that a post you admired for its human voice was actually written by a bot,"

Substack's blog says.

Relying on a third-party detector is a calculated risk. AI detectors are regularly criticized for false positives, especially on texts by authors for whom English is not a native language. That is why Substack leaves the final judgment to the reader instead of issuing a verdict on behalf of the platform.

What This Means

The fight over the "humanness" of content is moving from public debate into the product itself: platforms are starting to build AI detection directly into the interface. Substack is betting that subscribers want to read people, not models, and has given them a button to check. How accurate Pangram proves in practice will become clear in the feature's first months.

Frequently Asked Questions

What is Claudefishing?

Claudefishing is Substack's term for a situation where a text that impressed a reader with its "living" style turns out to have been written by an AI bot. The name is formed from Anthropic's Claude chatbot and the word "catfishing."

What does Substack's AI detection run on?

The feature is built on Pangram, a third-party service that specializes in recognizing AI text. Substack integrated it into its interface as an indicator button, but left the final decision on whether to trust a post to the reader.