Substack announced a new tool that helps readers estimate whether content on the platform was written or assisted by AI. According to a blog post published on Tuesday, the feature can scan posts, notes, replies, and comments and return an estimate of how much of the text may be AI-generated.
The detection capability is powered by a company called Pangram. Readers can analyze content longer than 100 words by selecting the “Scan for AI text” option from the three-dot menu in the top-right corner of a post. The tool is rolling out across the web and the iOS app, with an Android launch described as coming soon.
Why it matters
As AI-generated writing becomes more common, tools that flag possible machine authorship give readers a way to weigh what they encounter on the platform. The estimate applies to both fully AI-generated text and content written with AI assistance.
Who should care
Substack readers and writers on the platform are the primary audience, since the scanning option is available directly within posts and comments.