Imagine discovering that an AI tool is using your name and professional identity to sell a feature you never agreed to. That's exactly what happened to multiple journalists who found Grammarly had turned them into AI "expert reviewers" inside Superhuman's email client, attaching their real names to AI-generated writing suggestions sent to strangers.
The feature works like this: Superhuman's Grammarly integration offers an "expert review" function that rewrites users' emails. To give the AI suggestions more authority, it labels them with the names of real writers and editors, people who never signed up, never gave consent, and in many cases had no idea it was happening.
Opt-Out Is Not the Same as Consent
Grammarly's response has been to offer an opt-out process rather than switching to opt-in consent. That distinction matters. Opt-in means you choose to participate. Opt-out means a company uses your identity by default and puts the burden on you to discover it and ask them to stop.
This is a familiar playbook in tech. Companies scrape public data, build features on top of it, and frame the removal process as generous. But using someone's professional reputation to lend credibility to AI-generated text is a step beyond scraping blog posts for training data. It's borrowing trust that was never offered.
The affected parties include well-known tech journalists, but the scope appears to go much wider. Any published author whose work Grammarly has indexed could potentially show up as a named "expert" in someone else's inbox, endorsing AI suggestions they've never seen.
A Pattern Forming Across AI Companies
This sits in the same uncomfortable territory as AI art generators trained on artists' work without permission, or voice-cloning tools that replicate real people's voices. The common thread: AI companies treating individuals' creative identities as raw material, then asking forgiveness rather than permission.
What makes the Grammarly case particularly pointed is the directness of the identity use. Training on someone's writing style is abstract. Putting their name next to AI output and presenting it as their editorial judgment is concrete and personal. It's the difference between a company studying your recipes and a company putting your name on a restaurant you've never visited.
For anyone who uses writing tools with AI features, this is a reminder to check what permissions you've granted and what your name might be attached to. Grammarly has over 30 million daily active users. If you write professionally and have published work online, it's worth checking whether your name has been conscripted into someone else's product. The opt-out process, however it works, shouldn't be something you learn about from a news story months after the fact.