Grammarly has disabled its tool offering generative-AI feedback credited to real writers
TL;DR
Grammarly has disabled its 'Expert Review' feature, which delivered AI-generated writing feedback attributed to real authors, academics, and bloggers – without their knowledge or consent. The feature launched in August and relied on 'publicly available information from third-party LLMs,' widely understood as a euphemism for legally questionable web scraping. Both living and deceased public figures appeared as supposed feedback sources, ranging from bestselling novelists to tech bloggers.
Nauti's Take
'Publicly available information' is the AI industry's new favorite euphemism for 'we just took it. ' That Grammarly – a company whose entire brand is built on trust in writing quality – chose to misappropriate real people's names for AI output is not just legally risky, it's a reputational own-goal.
Quietly pulling the feature without an apology is the closest thing to an admission of wrongdoing that a PR team will ever allow.
Briefingshow
Grammarly's 'Expert Review' is a textbook case of AI companies using real identities as a trust proxy without legal grounding. Scraping public content to simulate living individuals touches both personality rights and copyright law. The feature's quiet removal – with no public apology – suggests Grammarly itself recognized the legal exposure it had created.
Sources