Grammarly has discontinued a controversial AI feature that suggested writing revisions modeled after well-known authors without their consent. The tool, which TechCrunch uncovered, offered "expert review" that essentially encouraged users to imitate the styles of prominent writers like Platformer's Casey Newton, who had no idea their voice was being used.

According to reports, the feature would pop up during writing to suggest changes that mimicked specific authors' tones and phrasing. Grammarly claimed it "fell short" on execution and removed the feature, but the incident highlights broader concerns about generative AI tools appropriating creative work. The company had apparently launched this without considering the ethical implications.

The episode underscores ongoing tensions in the writing community about AI's role in creative work. Many writers argue that tools like Grammarly's AI features, which can sound mechanical and repetitive, undermine the craft they've spent years developing. The controversy also raises questions about consent and attribution when AI systems train on or mimic living authors' distinctive voices without permission.