Grammarly is fighting a class action lawsuit after its AI-powered 'Expert Review' feature impersonated established authors and academics without permission. The writing platform abruptly killed the feature Wednesday, the same day legal papers hit, marking a major escalation in the growing backlash over AI tools cloning real people's voices and expertise. The case could set precedent for how AI companies use personal likeness and professional identity in their products.
Grammarly just became the latest AI company to face serious legal consequences for cloning real people without asking first. The writing assistant platform is now defending itself against a class action lawsuit challenging its 'Expert Review' feature, which presented AI-generated editing suggestions as if they came from established authors, professors, and writing experts—all without getting their permission.
The timing tells the story. Grammarly pulled the plug on Expert Review Wednesday, the exact day the lawsuit landed. That's not a coincidence. The feature had been live for months, offering users the chance to get feedback supposedly styled after recognized writing authorities. Except those authorities never agreed to have their names, reputations, or expertise borrowed by an algorithm.
According to the complaint, Grammarly's AI essentially created synthetic versions of real professionals, training on their work to mimic their editorial voices and then attaching their identities to machine-generated suggestions. Users paid premium prices thinking they were getting insights shaped by actual human experts. Instead, they got AI approximations trading on borrowed credibility.
This isn't just about hurt feelings or attribution disputes. The lawsuit argues Grammarly commercially exploited these individuals' professional reputations without compensation or consent. In the AI age, that's emerging as a crucial legal battleground. Your writing style, your analytical approach, your hard-earned expertise—can a company just feed that into a model and sell it back to customers under your name?
Grammarly hasn't issued a detailed public statement yet beyond confirming it discontinued the Expert Review feature. But the abrupt shutdown suggests the company recognized the legal exposure. The platform serves over 30 million daily users and counts major enterprises among its customers, making this a high-stakes fight that could influence how the entire productivity software industry approaches AI personalization.
The class action format means this could expand well beyond the initial plaintiffs. Any author, academic, or writing professional whose identity Grammarly used without permission could potentially join. That's a lot of people—the feature reportedly drew from dozens of recognizable names in publishing, journalism, and academia.
This case joins a growing wave of legal challenges to AI training practices and synthetic content. Earlier controversies around voice cloning and AI-generated impersonations have mostly hit entertainment and media companies. But Grammarly's troubles show the risk has spread to enterprise SaaS, where businesses are racing to add AI features without necessarily clearing the rights to the human expertise those features imitate.
The lawsuit's core argument is straightforward: you can't take someone's professional identity, run it through an AI model, and sell the output as if that person endorsed or created it. Grammarly's defense will likely hinge on whether AI-generated suggestions 'in the style of' someone constitutes impersonation or infringement—a question courts are just starting to tackle.
For Grammarly's competitors, this is a warning shot. Microsoft Editor, ProWritingAid, and other AI writing tools are all experimenting with personalized feedback features. How they attribute AI suggestions and what they promise users about expertise behind recommendations just became a lot more legally fraught.
The case also highlights the consent gap in AI development. Companies have been relatively cavalier about scraping public content for training data, arguing it's transformative use. But when you attach a real person's name to AI output and charge money for it, the legal calculus changes. That's not just training on publicly available work—it's commercial exploitation of individual identity.
Grammarly built its reputation on helping people write better. The irony is that Expert Review might have actually delivered useful suggestions. But delivering value doesn't excuse appropriating someone's professional identity without permission. The plaintiffs aren't arguing the AI didn't work—they're arguing Grammarly had no right to make it work using their names and reputations.
What happens next could reshape AI product development across the enterprise software landscape. If the plaintiffs win or Grammarly settles for significant money, expect every SaaS company with AI features to audit how they're using real people's expertise, voices, or identities. The era of 'move fast and ask forgiveness later' might be ending, replaced by actual consent requirements before you turn someone into training data.
Grammarly's legal trouble over Expert Review marks a critical moment for AI ethics in enterprise software. The lawsuit forces questions the industry has been dodging: Can companies monetize AI that impersonates real professionals without their consent? Where's the line between training on public data and commercial exploitation of individual identity? For writing tools, productivity platforms, and any AI product claiming expert-level performance, the answers will determine whether they can keep building features first and seeking permission never—or whether the consent economy is finally arriving for artificial intelligence. Grammarly's next move in court will signal how seriously the company takes those questions, and how much the rest of the AI industry should worry.
