Shishir Mehrotra’s Push to Remake Grammarly Shows the Risks of A.I. Leadership
Shishir Mehrotra’s push to turn Grammarly, now parent-branded as Superhuman, into an A.I. powerhouse has delivered a major backlash. In less than a year as CEO, he has rebranded the company, led acquisitions and pushed an aggressive pivot into A.I. agents. But one of those tools, an A.I. “Expert Review” feature launched last summer, has turned from showcase to liability. Expert Review offered users writing suggestions in the style of well-known authors and journalists, presenting feedback “from” figures who had never agreed to be involved. The tool, since disabled, has become a cautionary example of companies racing into A.I. without fully weighing reputational, legal and ethical risks.
Introduced in August, Expert Review promised tailored feedback from marquee names such as author Stephen King and astrophysicist Neil deGrasse Tyson. The catch was that none of the featured “experts” had consented to the project, which generated advice using A.I. models trained on their work and likenesses.
Writers quickly noticed and pushed back. “No one asked me for permission to use my name in this way, much less compensate me for whatever expert-reviewing labor my A.I. clone was apparently now doing on my behalf,” wrote journalist Casey Newton. In other cases, the feature invoked people who could not possibly have agreed. Vanessa Heggie, an associate professor at the University of Birmingham, called the appearance of historian David Abulafia, who died in January, “obscene.”
Earlier this month, the company was hit with a class action lawsuit alleging it used writers’ and journalists’ names for commercial purposes without consent, seeking more than $5 million in damages. “We have reviewed the lawsuit, and we believe the legal claims are without merit and will strongly defend against them,” Mehrotro said in a statement.
On March 11, the same day the suit was filed, Mehrotra posted a public apology on LinkedIn and announced that Expert Review would be paused. “We hear the feedback and recognize we fell short on this,” he wrote, describing the tool as part of Superhuman’s effort to bring A.I. directly to users. “I want to apologize and acknowledge that we’ll rethink our approach going forward.”
The controversy came about a year into Mehrotra’s tenure. He became CEO after Superhuman acquired Coda, the productivity startup he co-founded and led, at the end of 2024. Before that, he held senior roles at YouTube, GoogleTV and Microsoft. Since taking over, he has made clear that his ambition is to expand the company far beyond its roots as a grammar checker.
That strategy crystallized last October, when he announced that Grammarly’s parent company would be renamed Superhuman, while the writing tool would keep the Grammarly name. The rebrand was meant to signal a broader push into workplace productivity and email products as competition in A.I.-powered writing tools heats up.
Mehrotra has backed the vision with a rapid cadence of releases. Expert Review arrived as part of a bundle of A.I. agents promising to predict essay grades, surface relevant citations, simulate reader reactions and check for plagiarism. It is still unclear whether this ambitious pivot will pay off for Grammarly, which started 17 years ago as a grammar and spelling assistant and was last valued at $13 billion in 2021.
In response to the backlash, Mehrotra has suggested the feature will eventually return in a different form. He has said “there is a better approach to bringing experts onto our platform” and that Superhuman is working on a new version of Expert Review designed to “provide significantly more benefit to both users and experts.” Whether those experts will be willing to partner with the company after this episode remains an open question.