Grammarly Sued in Class Action Over AI Expert Identity Theft — Julia Angwin Files Lawsuit

Grammarly Sued in Class Action Over AI Expert Identity Theft — Julia Angwin Files Lawsuit

Grammarly is facing a landmark class action lawsuit over its AI-powered 'Expert Review' feature, which allegedly used the identities of real journalists and writers without consent to train and authenticate its AI suggestions. Filed on March 11, 2026, by Pulitzer Prize-nominated investigative journalist Julia Angwin, the complaint accuses the company of misappropriating personal writing styles, biographical details, and professional credentials to create the illusion of human expertise—without disclosure or authorization. This practice, according to court documents, violates consumer protection and privacy laws in multiple U.S. jurisdictions.

Background of the Allegations

According to Wired, Grammarly’s 'Expert Review' feature, marketed as offering insights from seasoned editors and writers, was trained using publicly available writing samples from journalists—including bylines, editorial histories, and tone patterns—culled from media outlets and personal blogs. The system then generated AI-driven suggestions that mimicked the voice and style of these individuals, often attributing advice to them by name in user interfaces. No consent was sought, and no compensation was offered.

How Grammarly’s AI Co-opted Real Experts Without Consent

The lawsuit, which seeks class-action status on behalf of over 500 identified individuals whose identities were used, claims this constitutes a form of digital identity theft under state consumer fraud statutes. Angwin’s legal team argues that Grammarly’s actions misled users into believing they were receiving feedback from real experts, thereby creating a false sense of authority and trustworthiness.

Grammarly’s Internal Communications and Privacy Gaps

While Grammarly has not issued a formal public statement, internal emails cited in the complaint reveal discussions among product developers about using "anonymized expert profiles"—a term that, according to the plaintiffs, was used to mask the use of identifiable individuals. Dataemia reports that the company’s privacy policy does not disclose the use of third-party identities in AI training, nor does it offer opt-out mechanisms for such usage.

Legal Precedents and AI Ethics Implications

The case has drawn scrutiny from AI ethics researchers and digital rights advocates. DNYUZ highlights that this is among the first lawsuits targeting generative AI systems for impersonating real people as part of their core functionality. Legal experts warn that if upheld, this precedent could force tech companies to overhaul how they source training data, particularly when human personas are involved.

What This Means for Users and AI Transparency

As the case moves toward discovery, affected journalists are being invited to join the class. For users, the implications are profound: if AI can convincingly impersonate real experts without consent, how can we trust any "expert" recommendation generated by machine learning? This lawsuit may become a defining moment in the future of AI ethics—and accountability.

How to Protect Yourself from AI Identity Theft

• Review the privacy policies of AI writing tools for data sourcing disclosures
• Opt out of data collection where possible, even if buried in settings
• Support legislation like the EU’s Digital Services Act and FTC AI guidelines
• Report suspicious AI-generated "expert" claims to digital rights organizations