Grammarly AI Lawsuit: Expert Review Risks Explained

The short version

Grammarly, the popular writing app owned by Superhuman, is facing a class action lawsuit because its AI "Expert Review" feature used the names and identities of real authors, journalists, and experts—like Stephen King, Neil deGrasse Tyson, and journalist Julia Angwin—without their permission to give writing advice. The feature pretended these people were reviewing users' text, but they didn't consent, and the advice was often bad or off-base. Grammarly has already shut down the feature due to backlash, and the lawsuit claims this breaks laws against using someone's name for profit without okaying it first.

What happened

Imagine you're writing an email in Grammarly, and it pops up with advice like, "Hey, Julia Angwin says make this sentence longer and fancier." Sounds helpful, right? But Julia Angwin—a real investigative journalist—didn't say that. She had no idea Grammarly was using her name, reputation, and style to power an AI tool that mimicked her as a virtual editor. The same went for big names like Stephen King or Neil deGrasse Tyson, and even everyday writers or professors. Grammarly's "Expert Review" feature, added last year as part of new AI tools, pulled from these people's public work to generate suggestions, with a tiny disclaimer saying they didn't endorse it.

Journalist Julia Angwin got mad when she found out (she learned about it from a tech newsletter) and filed the lawsuit in New York federal court on behalf of herself and "hundreds" of others. She called it like being "cloned" without permission, and pointed out the AI advice was often terrible—like suggesting to complicate simple sentences or add irrelevant ideas, making writing worse. Grammarly (run by Superhuman) faced huge backlash from writers on social media and in news outlets. Just before the lawsuit hit, they announced they'd disable the feature. Their product director, Ailian Gan, said they "missed the mark" and apologized, promising to redesign it so experts control if they're included. CEO Shishir Mehrotra echoed that on LinkedIn, admitting it misrepresented people's voices.

This isn't just drama—it's against laws in places like New York and California that say you can't use someone's name or likeness to make money without their okay. The suit wants damages over $5 million for the whole group and aims to stop this kind of thing.

Why should you care?

You probably use Grammarly (or something like it) to fix emails, essays, or work reports without sounding like a robot. This matters because it shows how AI tools are borrowing real people's brains and reputations to seem smarter, even if the advice stinks. If companies can fake "expert" input without asking, it erodes trust—next time Grammarly suggests something, will you wonder if it's really helpful or just AI pretending to be Stephen King? For everyday folks, it hits your wallet indirectly: lawsuits like this could make apps raise prices to cover legal bills, or force them to dumb down features. More broadly, it protects you—if you're not famous, your name or online writing could get scooped up tomorrow for some AI "helper" without your say-so.

What changes for you

Right now, nothing huge disrupts your Grammarly routine—the Expert Review feature is already turned off, so you won't see those fake celebrity critiques anymore. Your spelling checks, grammar fixes, and basic AI rewrites keep working as usual. But watch for updates: Grammarly says they're "reimagining" it to let real experts opt in or out, which could make future versions more trustworthy (or just remove the celeb gimmick).

If you're a writer, teacher, or anyone who's shared work online, double-check your privacy settings and think about how AI might mimic you. The lawsuit could set rules for all AI writing apps (like those in Google Docs or Microsoft Word), making them disclose exactly where advice comes from—no more sneaky impersonations. Prices might tick up slightly if legal fights drag on, but that's minor compared to gaining confidence that tools aren't lying about their smarts. And if you're in the "class" (anyone whose name was used), you might get notified later for possible payouts.

Frequently Asked Questions

### Is Grammarly still safe to use?

Yes, your core Grammarly experience—like grammar checks and basic rewrites—is unchanged and safe. They've disabled the problematic Expert Review feature, so no more unauthorized name-dropping. Just stick to the standard tools unless they roll out a fixed version.

### Who owns Grammarly, and why Superhuman?

Superhuman, a tech company, owns Grammarly now. They added the AI features last year to make it fancier, but this one backfired big time due to the consent issues.

### Can I get money from this lawsuit?

Maybe—if you're one of the affected people (like a writer whose name popped up), you could be part of the class action. Julia Angwin is the lead plaintiff, and damages are estimated over $5 million total. Keep an eye on news; courts will notify eligible folks if it moves forward.

### Why was the AI advice so bad?

The feature used AI (a large language model) trained on public writings, but it often garbled them into unhelpful suggestions—like overcomplicating sentences or adding off-topic fluff. Real experts like Angwin said it "made writing worse," showing AI isn't a perfect stand-in for human insight.

### Will this happen with other AI tools?

It could—this lawsuit tests laws on "right of publicity," so apps like ChatGPT or writing AIs might get stricter about crediting sources or getting permissions. Expect more disclaimers and opt-out options to avoid similar messes.

The bottom line

Grammarly got caught red-handed using real people's names to make their AI seem expert-level, sparking a lawsuit and forcing them to kill the feature— a win for creators whose identities can't be borrowed for profit without consent. For you, it means more trustworthy writing apps down the road, with less fake "celebrity" advice and clearer rules on AI tricks. It's a reminder to question AI claims, but your daily typing helper is fine for now. This could ripple to make all AI tools more ethical, protecting regular folks from digital cloning.

Grammarly Lawsuit Over AI 'Expert Review': What It Means for You