I second this. Get Substack Team to give us a newsletter-level opt-out right there in our Settings next to the AI scraper bot opt-out, instead of per article, and one major front of opposition to this Pangram integration will go away.

We’d also like direct confirmation that regardless of our opt-outs, none of our writing or inputs are being saved and used by Pangram for model improvement.

My communities’ other main concern, biases and inaccuracies for some important subgroups, remains. Looking forward to hearing what you are doing to mitigate those. You are still working on that, right? At Substack volumes, you’re looking at potentially thousands of false accusations a day.