'Obscene': Grammarly's New AI Tool Offers Writing Feedback From Dead Scholars
Grammarly’s new Expert Review AI feature provides writing feedback by simulating the perspectives of established scholars, journalists, and specialists, including some who are deceased. This has sparked criticism among academics, who argue that the tool effectively “resurrects” experts without their consent and uses their names and reputations to endorse feedback generated by large language models (LLMs). Grammarly (now branded Superhuman) says the tool draws on publicly available and widely cited works for inspiration but does not claim direct expert involvement. Users of the Superhuman Go extension can select from notable figures in relevant fields to receive guidance framed in their style. Critics say the lack of context, consent, or collaboration with the cited experts—especially those deceased—raises ethical concerns and risks eroding trust in AI educational tools. Similar trends are occurring with other companies, such as Meta and Khan Academy, which have also released AI programs that mimic real or historical individuals.
