When a writing app promises line‑by‑line feedback from famous journalists and authors, you might assume those experts actually agreed to help. That assumption may feel even safer when the app is the household name that is Grammarly, a San Francisco-based tech company that sells an AI-powered proofreading tool.
But now, the widely-used editing platform faces a class action lawsuit over its “Expert Review” feature, which allegedly used hundreds of real writers’ names and reputations to sell subscriptions without their consent.
‘Expert’ Help for $12 a Month?
For years, Grammarly has pitched itself as an AI writing assistant, with a free tier and paid upgrades that help users rewrite sentences, adjust tone, and detect plagiarism or AI‑generated text. According to the complaint, the company and its parent, Superhuman Platform, Inc., claim tens of millions of users and rapidly growing revenue to support an expanding product line.
One of those expansions was “Expert Review,” launched in August 2025 for paying users. For around $12 per month, the feature promises suggestions and comments on your writing from “experts” including bestselling author Stephen King, astrophysicist Neil deGrasse Tyson, New York Times tech reporter Kashmir Hill, journalist Kara Swisher, and former FTC commissioner Julie Brill.
How the Tool Uses Real Names
When subscribers turn on the Grammarly’s expert review feature, the complaint says Grammarly first tells them it is “reading your text” and “finding experts to review your piece.” Then, as the tool runs, it displays messages like “Applying ideas from Julia Angwin,” alongside a short bio describing her as an investigative journalist and author known for work on digital privacy and accountability.
Next to specific parts of a draft, the interface shows brief, attributed comments—for example: “Lead with personal stake,” “Sharpen the opening,” “Simplify technical tracker description,” or “Clarify the ‘so what’ after each finding.” If users click on a comment, they see a longer suggestion, which Grammarly describes as “inspired” by that particular expert, and can insert an example revision with a button click.
According to the lawsuit, this design is no accident. It alleges the feature is structured so users believe that named experts like Angwin are actually providing their perspectives and feedback, or at least are directly associated with the advice on screen. For fans ChatGPT or other chatbots, the whole setup will feel familiar: an AI layer offering confident, personalized suggestions — except here it’s doing so under real journalists’ names they never agreed to lend.
Journalist Says She Never Signed Up
Lead plaintiff Julia Angwin is a New York–based investigative journalist and editor, currently a New York Times contributing Opinion editor and founder of the nonprofit newsroom The Markup. Her resume includes stints at The Markup, ProPublica, and The Wall Street Journal, as well as a Pulitzer‑winning WSJ team and multiple Loeb Awards, plus two books on surveillance and technology.
Angwin alleges she never agreed to serve as a Grammarly “expert,” never licensed her name or identity for the AI feature, and never approved any of the comments attributed to her. Instead, the complaint says, Grammarly scraped her publicly available work, used it to train its systems, and then generated “advice” that it attributed to her — even where she might disagree with the recommendations or would give different guidance depending on context.
She says she only learned about the feature from a March 2026 article by another journalist, who discovered that he and others had been “involuntarily conscripted into serving as unpaid experts.” Around the same time, tech leaders like Shishir Mehrotra were posting on LinkedIn about the risks of AI‑driven “expert” overlays that borrow real people’s reputations without clear consent or disclaimers, underscoring how uneasy many subject‑matter experts are with this kind of design.
Angwin describes being “shocked and horrified” that an AI product was using her name for profit without her consent, warning that users could blame her personally if they followed “her” advice and got a bad grade, poor evaluation, or other negative outcome, even though she had nothing to do with it.
Right of Publicity
The suit casts Grammarly’s conduct as a right‑of‑publicity case for the artificial intelligence era, with Angwin suing its owner, Superhuman Platform, Inc., in Manhattan federal court on behalf of a nationwide class of “experts” whose names were allegedly used without consent. It leans on New York and California right‑of‑publicity laws that bar using someone’s name or likeness for commercial gain without consent, plus California’s common‑law protection for the economic value of a person’s identity.
The complaint argues that by placing real journalists’ and authors’ names in its interface, telling users it is “applying ideas” from those people, and selling subscriptions on that basis, Superhuman appropriated their identities for commercial purposes in violation of both states’ laws. A separate unjust‑enrichment claim alleges the company profited from those names and reputations without sharing any of the resulting revenue with the people whose identities it traded on.
Proposed Class, Harm, and the Stakes
Angwin seeks to represent a nationwide class, plus a New York subclass, of hundreds of “experts” whose names were used in Expert Review without their consent. The complaint says the core harm is losing control over how those names sell a product without compensation.
The case lands amid broader debates over how far AI tools built on large language models (LLMs), including chat‑GPT‑style ai agents from companies like OpenAI, can use real people’s work and personas. On paper it’s a classic right‑of‑publicity suit, but it also asks how close an AI that claims to “channel” a human expert comes to putting words in their mouth, especially when the comments themselves are AI-generated.
Related Resources:
- Is Grammarly Generative AI? If So, Do Lawyers Need to Disclose Its Use? (FindLaw’s Practice of Law)
- Appropriation of Name and Likeness (FindLaw’s Learn About the Law)
- New Lawsuit Claims ChatGPT Practices Law Without a License (FindLaw’s Practice of Law)
