Superhuman used journalists' names without permission — CEO explains on air

Grammarly, rebranded as Superhuman, launched an Expert Review feature that generated editorial advice attributed to real journalists and experts—without their knowledge or consent. Among the "experts" featured were The Verge editor Nilay Patel and investigative journalist Julia Angwin. The scandal escalated so dramatically that Angwin filed a class action lawsuit, and Superhuman CEO Shishir Mehrotra agreed to an interview with Patel—the very person whose name his company used without permission.

Superhuman is an American AI company that rebranded from Grammarly at the end of 2025. The flagship Grammarly product is familiar to approximately 40 million daily users as an AI writing assistant. The company is also developing Coda (documents), Mail email client, and the Superhuman Go platform—a suite of personalized AI agents that work directly within the applications employees already use.

The Expert Review feature launched in August 2025. It offered users editorial corrections "inspired by" specific media figures—with mentions of their names and links to their publications. No one was asked for permission.

During the interview, Mehrotra acknowledged that the feature was a failure—both for users and experts. It was barely used, didn't align with the company's strategy, and was already disabled before the lawsuit was filed. However, Superhuman's first response was an email unsubscribe option, not immediate removal.

The sharpest moment of the interview became a discussion about attribution and impersonation. Mehrotra insisted that Expert Review was not engaged in impersonation: the feature merely "inspired" itself from the work of specific people and clearly indicated this. Patel countered that what the system generated in his name had nothing to do with his actual editorial advice.

"I never talk about emotional headlines. Vergecast is not a show about editing headlines about smartwatches." According to him, attributing someone else's text is not the same as putting someone else's name under generated content that the person would never have created.

Angwin's lawsuit is based not on impersonation but on New York and California laws prohibiting the use of people's names and likenesses for commercial purposes without consent. Mehrotra insists the claims have no merit and notes that no LLM today receives attribution permission. However, he acknowledges that the legal minimum is not the standard to strive for.

His benchmark is the YouTube model, which he himself built: a platform that creators would want to join of their own accord, with a clear revenue share. Superhuman Go plans to share 70% of revenue with agent creators. The conversation extended beyond a single incident.

Patel cited an NBC News poll showing that public perception of AI is worse than that of immigration services. In his view, the reason is the extractive nature of the technology: AI takes others' labor as raw material without offering anything of comparable value in return. Mehrotra disagreed: people fear job loss, not attribution problems.

According to him, Superhuman's goal is not to replace people but to augment their capabilities, to turn them into "superhuman." This episode is a symptom of a systemic contradiction. AI companies build products on human creativity without establishing a clear economic agreement with authors.

Courts have yet to answer the key question: is training LLMs on others' texts a violation of copyright. As long as this question remains open, such scandals are inevitable. Superhuman closed the feature and apologized—but the lesson for the entire industry is far broader.