Grammarly Drops AI Feature That Impersonated Writers Without Consent in 2026

Grammarly Drops AI Feature That Impersonated Writers Without Consent in 2026

Grammarly has permanently removed its "Expert Review" AI feature after intense backlash for impersonating professional writers without consent. Launched in early 2026, the tool used public writing samples to mimic the voices of renowned authors and editors—misleading users into believing they received feedback from real experts like Kara Swisher. Critically, users and authors were never informed their content was used to train the system.

How the Expert Review Feature Worked

The AI analyzed millions of publicly available writing samples, including articles, books, and social media posts, to replicate distinct stylistic patterns. It then generated edits that mirrored the tone and phrasing of identified "experts," creating the illusion of personalized feedback from top editors. No disclosure was provided to users or original authors.

User Backlash and Media Coverage

Investigative journalist Casey Newton of Platformer exposed the feature, revealing he was impersonated without permission. His report sparked outrage among writers and ethicists. Users on The Write Practice and Reddit reported receiving contradictory edits that distorted their voice. The term "AI style mimicry" trended across writing communities as trust in AI editing tools plummeted.

Grammarly’s Official Statement and Response

On March 9, 2026, Grammarly issued a public apology: "We hear the feedback and recognize we fell short on this." The company confirmed immediate removal of the feature and pledged to implement explicit consent protocols for future AI training involving human writing. However, it did not disclose whether user data was deleted or how training datasets will be audited.

Broader Implications for AI Ethics

This incident highlights a growing crisis in AI ethics: unauthorized replication of human creativity. Experts at Binance Square warn that "social media and product growth must be underpinned by explicit consent." Regulatory bodies are now scrutinizing AI content generation, particularly around identity appropriation and intellectual property rights. The incident may accelerate legislation like the EU AI Act’s proposed rules on synthetic identity.

What This Means for AI Writing Assistants

As AI tools become more sophisticated, the line between augmentation and impersonation grows dangerously thin. Grammarly’s core grammar-checking tools remain unaffected, but the removal of Expert Review signals a retreat from unchecked experimentation. Users now demand transparency: Is the editor human? Is the AI disclosed? Is my data used ethically? Without answers, trust erodes.