Substack grapples with AI tools and extremist content amid growing writer base

In early June, bestselling author Naomi Klein announced a Substack, presenting it as a quieter outlet compared with Twitter and Instagram, yet many online commentators condemned her for endorsing a service accused of facilitating white-supremacist, Holocaust-denial and neo-Nazi newsletters. A February Guardian probe found Substack not only takes a cut of subscription fees for such content but also boosts it through its algorithms.

As AI-generated essays proliferate, Substack rolled out an AI-detection feature named Pangram, which some observers praise for transparency but criticize as a distraction from extremist moderation. Platform co-founder Hamish McKenzie and head of Canadian partnerships Mark Swierszcz defend the hands-off policy, arguing that demonetising or removing content would worsen the problem. Critics like tech reporter Casey Newton note that growth incentives may soon lead to ad-supported free newsletters, while writers such as Danielle French continue to value Substack’s simple publishing tools for local news. The debate underscores tension between the venture-backed expansion and Substack’s original promise of a clean, direct writer-to-reader channel.