Writing Tool Removes AI Personas After Legal Challenge from Authors

Grammarly has disabled an AI feature that mimicked the writing styles of prominent authors and scientists without their consent, following a significant legal challenge from the writers whose identities were used. The Expert Review function, which offered writing feedback “inspired by” the personas of figures including Stephen King and Carl Sagan, was taken down this week by Superhuman, the tech firm that operates Grammarly. The move came after a multi-million dollar lawsuit was filed in the Southern District of New York by investigative journalist Julia Angwin and other writers who discovered their names and professional reputations being marketed as commercial AI personas. Superhuman’s chief executive acknowledged the tool had “misrepresented” the voices of experts|expert voices, apologizing for the controversial feature.

The Feature That Triggered Controversy

The Expert Review capability represented a significant shift from Grammarly’s established editing support approach. Rather than delivering one-size-fits-all feedback, the tool enabled users to get editing suggestions “inspired by” the distinctive voices of famous writers and academics. Users could select from personas including acclaimed novelist Stephen King and renowned scientist Carl Sagan, among numerous other public figures. The feature promised to offer personalized writing advice examined through the lens of these distinguished experts, ostensibly assisting users improve their work by drawing inspiration from the best in their respective fields.

What Grammarly presented as an innovative educational tool soon became apparent as a problematic appropriation of identity and intellectual property. The company had not sought permission from any of the writers whose personas were being copied and sold. Reporter Julia Angwin, who became the lead plaintiff in the class action suit, expressed shock at discovering her professional identity being marketed as a product feature. She characterized the matter as fundamentally different from traditional deepfakes, stressing that her editing abilities constitutes her income and that she had never imagined her professional skills could be taken and repackaged this way.

• AI personas replicated hundreds of author voices without permission or payment
• Feature promised feedback inspired by famous authors and scientists
• Users could select different expert personas for revision feedback
• Tool was incorporated into Grammarly’s paid subscription service offerings

Legal Action and Industry Response

The legal challenge against Superhuman and Grammarly constitutes a significant moment in the wider conversation over AI ethics and IP protections. Led by investigative journalist Julia Angwin, the class-action legal action filed in the Southern District of New York claims that the company wrongfully used the names of hundreds of writers to generate profits from its paid service. The filing maintains that utilizing identities and professional reputations for commercial purposes without explicit consent breaches existing legal protections safeguarding people from unlawful commercial use of their identity and image.

The reaction to the lawsuit has been swift and substantial. Within 24 hours of filing, Angwin’s legal team reported hearing from over 40 prospective claimants keen to participate in the action, demonstrating broad apprehension among impacted authors. The case pursues damages exceeding $5 million, though industry analysts suggest the real amount could be considerably higher once the court determines compensation based on the company’s revenue generated by the disputed feature. Superhuman’s rapid decision to disable the Expert Review function indicates the company recognized the reputational and legal risks created by maintaining the feature.

The Court Case Details

The lawsuit specifically contends that Grammarly and Superhuman violated fundamental principles of personal identity safeguards by crediting writing guidance to authors that failed to give such advice. The lawsuit filing emphasizes that the firm profited from these personas through its paid membership structure, creating profits directly from the unlicensed use of numerous individuals’ identities and career standing. Attorneys contend this represents a “blatant violation of the law,” pointing to established court rulings safeguarding people from unauthorized commercialization of their personal identity without consent.

Julia Angwin’s direct frustration with the feature surpassed the regulatory breaches to the standard of the AI-generated content. She labeled the editing suggestions assigned to her as a “slopperganger”—a term referring to substandard machine-created text—observing that the edits were making sentences more problematic rather than enhancing them. This added dimension to the case highlights not only the regulatory violation but also the damage to professional standing of having one’s identity linked to inferior quality output, magnifying the harm of improper use of one’s name.

• Damages sought surpass $5 million with final amount based on company earnings
• More than 40 other claimants reached out to legal team in the initial day
• Claims unlawful commercial use of identities lacking consent or compensation

Reliability Problems and Credibility Issues

Beyond the legal violations, the Expert Review function prompted significant concerns about the reliability and accuracy of algorithmically-produced editorial guidance. Users depending on suggestions credited to recognized authors and scholars had no means to determine whether they were receiving genuine guidance or computer-generated imitations of professional expertise. This erosion of trust extends beyond single users to the wider writing sector, where audiences and learners might have reasonably assumed they were receiving instruction from established authorities. The feature’s removal highlights a critical gap between what artificial intelligence can technically achieve and what it ought to be allowed to do from an ethical standpoint.

The reputational harm experienced by impersonated writers was especially insidious because it connected their identities with low-quality output. Angwin’s case illustrated the issue—her professional identity was being promoted as a top-tier service while at the same time providing substandard editorial support. This mix of misuse of identity and low quality caused dual harm: loss of control over her name paired with linking to substandard results that contradicted her professional expectations. For writers whose credibility is built on the standard of their work, such misrepresentation poses an existential threat to their reputation and commercial value.

The Challenge with AI Mimicry

The core issue in Grammarly’s methodology lay in attempting to replicate the subtle discernment and skill of seasoned professionals through automated systems. Skilled revision requires contextual understanding, attention to style, and years of accumulated experience—elements that cannot genuinely be replicated by studying written work and producing edits in a similar voice. Angwin’s point that the machine-produced revisions made sentences increasingly complicated rather than making them better demonstrated the fundamental emptiness of the mimicry. The system could replicate superficial style elements but lacked the profound understanding necessary to provide authentically helpful direction, ultimately weakening both the genuineness of the voices and the usefulness of the platform itself.

Organizational Response and Next Steps

Superhuman’s chief executive Shishir Mehrotra acknowledged the misstep publicly, releasing an apology on LinkedIn in which he admitted that the Expert Review function had “misrepresented” the voices of the impersonated experts. The company’s swift decision to turn off the feature this week indicates an attempt to mitigate additional legal and reputational damage. However, the deactivation came only following the lawsuit was filed and substantial public criticism surfaced, prompting concerns about whether the company would have taken action without external pressure. Mehrotra’s statement, while apologetic in nature, did not tackle the larger issue of how such a feature was approved and deployed in the first place, nor did it outline specific steps to avoid similar incidents in the future.

The road ahead for Grammarly remains uncertain as the litigation proceeds. Beyond the urgent legal dispute, the company faces the challenge of rebuilding trust with writers and users alike who may now scrutinize the moral principles governing its AI development. The elimination of the Expert Review function demonstrates a defensive rather than forward-thinking stance, suggesting the company is responding to legal pressure rather than exhibiting authentic dedication to ethical artificial intelligence. Moving forward, Grammarly will likely need to introduce more stringent approval procedures and oversight mechanisms for any capabilities that employ the utilization of individuals’ names or images. The company’s response to this situation may set precedent for how rival artificial intelligence companies address the use of well-known individuals in their AI tools.

The speed at which Grammarly disabled the feature points to the company acknowledged the legal, reputational, and business stakes involved. However, the lack of proactive measures before the lawsuit reveals that review procedures missed the ethical violations. As the legal proceedings continue, the company could encounter additional scrutiny regarding how numerous content creators were impacted and whether damages will be provided more than what the lawsuit requires. The case is poised to affect how rival AI organizations handle the inclusion of real identities in their products in the future.