AI Writing Tool Misuses Expert Names Without Consent

Grammarly's "expert review" feature, launched in August, has drawn criticism for using AI to generate writing advice inspired by subject matter experts. The feature analyzes user text and provides suggestions, allegedly drawing from influential figures like Stephen King and Carl Sagan. However, a significant issue has arisen as numerous journalists and prominent writers, including those from The Verge, Wired, and Bloomberg, have found themselves listed as "experts" without granting permission.

Some experts noted inaccuracies in their descriptions, such as outdated job titles, which could have been corrected had their input been sought. In response, Grammarly's parent company, Superhuman, stated that the feature provides suggestions inspired by publicly available works and does not claim endorsement. The company also mentioned that experts are included because their published works are widely cited and publicly accessible.

Further complicating matters, users attempting to "explore more deeply" the sources of these AI suggestions encountered frequent crashes and found links directing to spammy websites or unrelated content. This raises concerns about the accuracy and reliability of the AI's recommendations, with some sources potentially misattributed, suggesting the AI might base advice on one person's work while citing another.

The presentation of these AI suggestions also raises ethical questions. In platforms like Google Docs, the suggestions closely resemble real user comments, creating a misleading impression of direct expert feedback. This mimics the experience of receiving edits from a human expert, potentially leading users to accept AI-generated advice that misinterprets an individual's preferred editing style.