Grammarly has recently withdrawn a contentious feature that utilized artificial intelligence to generate written feedback resembling the insights of both contemporary and historical academics, all without obtaining their permission. This move comes as the company confronts a class action lawsuit.
The widely-used grammar checking software, which is endorsed by numerous universities, has encountered significant criticism related to its “expert review” feature that was introduced last year.
This tool enabled subscribers to input their written work into the software and receive AI-generated critiques purportedly from “subject matter experts.” These experts ranged from renowned authors like Stephen King and Neil deGrasse Tyson to lesser-known academics. Alarmingly, none of those named seem to have been consulted beforehand.
A legal complaint, submitted by journalist Julia Angwin on March 11, contends that Superhuman, the parent company of Grammarly, has compromised her privacy by utilizing her identity without consent, thereby interfering with her ability to control the commercial use of her name.
According to the lawsuit, “In this action, Julia Angwin, an award-winning journalist and editor, challenges Grammarly’s misappropriation of the names and identities of hundreds of journalists, authors, writers, and editors to generate profits for Grammarly and its owner, Superhuman.” This was initially reported by Wired.
Times Higher Education has reached out to Grammarly for a response.
On the same day as the complaint, Superhuman announced they would disable the expert review feature due to the critical feedback received. “The agent draws on publicly available information from third-party large language models to offer writing suggestions inspired by the published work of influential voices,” CEO Shishir Mehrotra stated in a LinkedIn post. “Recently, we’ve received valid critical feedback from experts who are concerned that the agent misrepresented their voices.”
In the UK and US, many universities subscribe to Grammarly and actively encourage students to use it. On its website, Grammarly lists notable institutions such as Stanford University, Boston University, and the University of Pennsylvania among its numerous clients.
The University of Birmingham has previously released guidelines on using Grammarly, although these do not specifically address the expert review feature, and it remains unclear when these guidelines were issued. Times Higher Education has contacted the university for further details.
John Kaag, a philosophy professor at the University of Massachusetts Lowell, remarked that the presence of scholars in the Grammarly tool “implies endorsement or participation that was never agreed to.”
In a contrasting approach, Kaag launched his own chatbot for the company Rebind after dedicating 30 hours recording his thoughts and answers to questions regarding Henry David Thoreau’s memoir Walden, which was then used to train an AI based on his actual words.
“In Rebind, we contract original commentary about classic books, generating hours of dialogue from some of the world’s leading educators,” he explained. This setup facilitates a guided conversation about literature, drawing directly from his authentic commentary with appropriate attribution.
Kaag criticized Grammarly’s position, noting that “the expert review does not claim direct participation from those experts – the suggestions are described as inspired by their work rather than produced by them. This is just garbage.” He argued that Grammarly exploits recognized scholarly authority as a design choice to enhance the perceived credibility and specificity of AI feedback.
