Substack's new tool tells you who's been writing their newsletters with AI
23 July 2026 · filed under 15ce799930d9
Substack has introduced a tool that estimates how much of a given newsletter was written with the help of AI, according to TechCrunch. The feature gives readers a rough measure of AI involvement in a piece of writing, rather than a simple yes-or-no verdict on authorship.
The move fits a broader industry pattern in which platforms are adding disclosure mechanisms for AI-assisted content rather than banning such content outright. TechCrunch frames the tool as a signal of Substack’s stance: AI use in newsletters is treated as a matter for transparency rather than prohibition.
The report does not detail the underlying detection method, nor does it specify how the estimate is calculated or displayed to readers. It is described as a step toward giving audiences more information about how the content they read was produced.
TechCrunch notes that the change carries implications beyond Substack’s own platform, touching on questions of trust between writers and readers, potential interest from regulators watching AI disclosure practices generally, and the economics of newsletter writing, where AI assistance could affect how subscribers value and pay for a writer’s work.
No statement from Substack executives is quoted in the report, and no timeline for rollout to all users is specified. The tool’s introduction is presented as part of an ongoing conversation about how platforms handle AI-generated and AI-assisted text.