Grammarly Pulls AI Feature That Impersonated Writers Without Their Consent
Grammarly has disabled an AI feature that mimicked the editorial voices of famous writers and academics including Stephen King, the late Carl Sagan and investigative journalist Julia Angwin after a fierce backlash and a class-action lawsuit accusing the company of stealing professional identities for commercial gain.
The feature, called Expert Review, was available to subscribers paying $144 a year and offered AI-generated writing feedback presented as though it came from prominent figures in journalism, science and literature. The named experts were not asked for permission, and many only discovered their names were being used when the story became public.
Grammarly's parent company Superhuman said the tool drew on "publicly available information from third-party LLMs to surface writing suggestions inspired by the published work of influential voices", which critics have read as a tacit admission that web scraping of uncertain legality was involved. The company included a disclaimer on the service saying the references "do not indicate any affiliation with Grammarly or endorsement by those individuals."
Angwin, a New York Times contributing opinion writer who has spent her career investigating tech companies' impacts on privacy, said she was "stunned" to find her professional identity being sold as a product. "I had thought of deepfakes as something that happens to celebrities, mostly around images," she told the BBC. "Editing is a skill... it's my livelihood, but it's not something I've ever thought about anyone trying to steal from me before. I didn't even think it was stealable."
She also took a dim view of the quality of the imitation, calling it a "slopperganger." "The edits were not good. The ones that they were attributing to me... were making the sentences worse, more complex," she said. "The idea that my name would be in there giving people terrible advice is actually really appalling."
The lawsuit, filed in the Southern District of New York, alleges that the company misappropriated the identities of "hundreds" of writers to drive subscriptions. Angwin's lawyer Peter Romer-Friedman said they had "heard from over 40 people in the last 24 hours since we filed the suit," calling the company's conduct a "brazen violation of the law." While the filing states damages exceed $5m, Romer-Friedman noted this is a minimum jurisdictional threshold, and that the final figure will depend on what the company earned from the tool.
The quality of the AI's output did little to help Superhuman's case. Tech journalist Casey Newton fed one of his own articles into the tool and received feedback from Grammarly's approximation of Kara Swisher, which provided feedback that was so generic that it prompted him to relay it to the actual Swisher. Her response was unambiguous: "You rapacious information and identity thieves better get ready for me to go full McConaughey on you. Also, you suck."
When the backlash over the feature first started, Superhuman initially said it would keep the feature running but allow named experts to opt out by email, which gaming journalist Wes Fenlon, who is one of the people the AI is impersonating, called "a laughably inadequate recourse for selling a product that verges on impersonation and profits on unearned credibility." Romer-Friedman argued that the burden of consent should never have been placed on the writers to begin with.
Superhuman CEO Shishir Mehrotra has since apologised on LinkedIn, writing that the company had "received valid critical feedback from experts who are concerned that the agent misrepresented their voices" and that it had "fallen short." He confirmed that Expert Review had been pulled for a redesign and said that the firm was working on a "better approach to bringing experts onto our platform" that would "benefit both users and experts."
Responding to the lawsuit specifically, Mehrotra told the BBC the feature had been taken down before the claim was filed and had seen "very little usage" in its short existence. He apologised but maintained the legal claims were "without merit" and said that the company would "strongly defend against them."
Grammarly was founded in 2009 as a writing review tool and began integrating generative AI in August 2025. The company began rebranding to Superhuman in October but kept Grammarly as the name of its core service.
