Key Takeaways
- Substack integrates AI detection software to identify AI-written content.
- The tool aims to enhance transparency for readers regarding authorship.
- Writers can disclose AI usage in their posts.
- Publishers can review their drafts for AI content before publication.
New AI Detection Feature
Substack has rolled out a new feature that allows users to determine which newsletters are partially or fully written by AI. This integration with Pangram, an AI writing detection tool, enables readers to assess the human versus AI contribution in posts, comments, and replies on the platform.
Potential Impact on Trust
While this feature may initially pose challenges for Substack, as it could reveal that some newsletters are not entirely human-written, it also has the potential to strengthen trust in the platform. By providing clarity on authorship, Substack aims to maintain its reputation for high-quality content.
Industry Trends
Substack joins a growing number of platforms that are beginning to label AI-generated content. Social media sites have already implemented similar measures for AI-generated images and videos, while music streaming services are starting to identify AI-created music.
Writers and Publishers Empowered
Substack’s CEO, Chris Best, emphasized the importance of human creativity in writing. He noted that while AI can assist with many tasks, the core idea and message should come from the writer. The new feature will be available for any content over 100 characters, allowing writers to include an optional note about their use of AI.
Publishers can also utilize Pangram on their drafts before they go live, ensuring they have control over how AI is represented in their work. This approach encourages transparency without penalizing the use of AI in writing.