Editorial illustration for Grammarly faces class-action suit over AI ‘Expert Review’ feature
Grammarly Sued Over Misleading AI 'Expert Review' Claims
Grammarly faces class-action suit over AI ‘Expert Review’ feature
Grammarly promised a ghostwriter. Instead, a group of authors say it built a ghost. A new class-action lawsuit, filed this month, targets the company’s “Expert Review” feature.
That tool claimed to let users receive feedback from specific veteran writers, living or dead. Not a single one was consulted. The suit alleges Grammarly used their names, reputations, and life’s work to train its AI.
Customers paid for an expert’s critique. What they bought was a simulation, constructed from a pilfered identity.
We are sorry and will do things differently going forward." As WIRED reported earlier this month, Superhuman last year added a suite of AI-powered widgets to the platform, including one that purported to have a veteran writer (living or dead) weigh in with a critique of the user's text. While a disclaimer clarified that none of the people cited had endorsed or directly participated in the development of this tool, which leveraged an underlying large language model, various writers, including WIRED journalists, expressed frustration over Grammarly invoking their likenesses and apparently regurgitating their life's work with these AI agents.
That apology is a legal maneuver. It doesn’t undo the act. The lawsuit, spearheaded by journalist Julia Angwin, makes a blunt argument: a disclaimer is not a license.
For the writers, it was theft. Their style became training data. Their name sold the product.
This case now pushes into a raw legal frontier. It asks how much of a professional identity a company can strip-mine and repackage before the transaction becomes fraud. Grammarly says it will change.
A court will decide what that change is worth.
Related Reading
Policy & Regulation Claude Mythos highlights EU AI safety gaps, says researcher Caroli
Policy & Regulation Data Center Watch Unfunded, Separate from 10a Labs' AI Risk Services
Policy & Regulation After ditching AI fitness apps and a Fitbit, I return to Peloton classes
From Archives Anthropic files lawsuit against U.S. government in federal court
From Archives OpenAI, Google staff file amicus brief for Anthropic vs Pentagon blacklist
Common Questions Answered
What specific claims does the class-action lawsuit against Grammarly make about its 'Expert Review' feature?
The lawsuit alleges that Grammarly misleads users by presenting AI-generated feedback as if it comes from a 'veteran writer' who does not actually exist or review the text. The complaint suggests that this marketing approach exploits the reputations of writers and scholars without their permission or consent.
Who is the named plaintiff in the lawsuit against Grammarly, and what is her primary argument?
Julia Angwin is the named plaintiff in the class-action lawsuit against Grammarly. She argues that the company's practice of framing AI suggestions as coming from well-known writers and scholars is deceptive and inappropriately uses these individuals' reputations without their agreement.
How does the Grammarly lawsuit relate to a similar incident involving Superhuman's AI writing tools?
The Grammarly lawsuit follows a similar pattern to Superhuman's AI writing tool controversy from the previous year, where the email client added AI widgets that claimed to provide critiques from veteran writers. Both cases involve AI-generated feedback being marketed as coming from specific, recognizable writers who never actually endorsed or participated in the tool's development.
Further Reading
- Grammarly's 'expert review' is just missing the actual experts — TechCrunch
- Grammarly Offering Manuscript Reviews by AI Versions of Recently Deceased Professors — Futurism
- Grammarly turned me into an AI editor against my will and I hate it — Platformer
- Grammarly's AI Expert Review Sparks Backlash Over Consent — eWeek
