Grammarly’s ’Expert Review’ Draws Criticism for Using Deceased Scholars’ Names

Grammarly (rebranded under parent Superhuman) launched an AI feature called Expert Review that provides writing feedback framed through the perspectives of named experts — including journalists, academics and some deceased scholars. The feature uses the company’s large language models to surface suggestions “inspired by” publicly available and widely cited works and lets users opt into expert-styled critiques via the Superhuman Go browser extension. Academics have complained the tool appears to ‘resurrect’ dead scholars without consent, calling the practice “obscene” and ethically troubling, while others warn it could deepen distrust of AI in education. Grammarly says Expert Review doesn’t claim endorsement from the named experts and that selections are based on publicly available publications. The rollout follows similar industry moves to create persona-driven AI assistants (e.g., Meta’s celebrity chatbots, Khan Academy’s historical role-play tutors). Key names highlighted by tests include Margaret Sullivan, Jack Shafer, Lawrence Lessig, Timnit Gebru and Helen Nissenbaum. The controversy centers on consent, attribution, and academic trust rather than any direct technical risk to crypto markets, but it underscores broader regulatory and reputational risks for AI product makers.