Substack’s new tool detects AI-written newsletters
▼ Summary
– Substack has integrated Pangram AI detection software to estimate how much of a post, comment, or reply was written by AI versus a human.
– In the short term, this feature may damage trust in Substack by revealing many newsletters are not entirely human-written.
– In the long term, AI detection could reduce “AI slop” and increase reader trust in human-created content.
– Substack joins other platforms in labeling AI content, as AI’s role in content creation grows.
– The feature is available for content over 100 characters and allows writers to add optional AI author’s notes, without penalizing AI-assisted writing.
Substack has introduced a tool that lets readers see how much of their favorite newsletters is actually written by a person. The platform announced a new integration with AI writing detection software Pangram, enabling users to scan posts, comments, and replies within the Substack app for an estimate of human versus AI-generated content.
In the near term, this move could hurt Substack’s business by revealing how many newsletters on the platform are not entirely human-crafted. That might erode trust in its ecosystem of independent journalism and blogs, or damage its reputation for hosting high-quality content. Over time, though, AI-detection features could help Substack stay free of “AI slop” and encourage readers to trust what they’re reading came from a person, or at least understand when it didn’t.
Substack joins a growing list of platforms leaning toward labeling AI-generated content. Social media sites already tag AI-made photos and videos, and music streaming services have begun labeling and sometimes penalizing AI-created music. “This is good use of AI,” said Substack CEO Chris Best. He recalled pitching the platform to writers by saying, “We’ll do everything for you except the hard part. You have to have an idea that’s worth reading, worth caring about, worth sharing. That one thing is very hard and very valuable. Software should do everything else, but you do want the person to do the hard part.”
The feature will be available in the Substack app for any post, note, reply, or comment longer than 100 characters. Substack also plans to let writers include an optional AI author’s note to properly disclose their use of AI, the company told TechCrunch. The tool is not designed to prohibit or penalize AI-assisted writing, but rather to encourage creators to add a “how I make this” statement explaining their process. Publishers can run Pangram on their own drafts before publication and report or remove scans they believe are inaccurate on their own work.
(Source: TechCrunch)
