Substack adds AI detection tool Pangram to fight fake human content

Substack, a platform for independent creators, has added an AI detection tool called Pangram to help readers see if content is written by humans or AI. The platform says the real issue isn't AI itself, but AI content disguised as human work, which they call 'Claudefishing'—readers expect real human insights but end up wasting time on content without human thought.

To boost transparency, Substack also launched tools like process notes and pre-publish Pangram checks. It plans to add AI preferences for replies and recommended content settings to improve user experience.