Grammarly uses our identities without permission

Grammarly’s Expert Review feature offers users writing advice “inspired by” subject matter experts, including recently deceased professors. Wired I mentioned Wednesday. When I tried this feature myself, I found some experts who came as a surprise for a different reason — one of whom was my boss.

The AI-generated comments included comments that appeared to be from EdgeThe magazine’s editor-in-chief, Nilay Patel, as well as editor-at-large David Pearce and senior editors Sean Hollister and Tom Warren, none of whom have given Grammarly permission to be included in the magazine. “Expert Reviews.”

Feature that It was launched in Augustclaims to help you “refine your message through the lens of relevant industry perspectives.” When users select the “Expert Review” button in Grammarly’s sidebar, it analyzes their writing and displays AI-generated suggestions “inspired by” relevant experts. “Industry Perspectives” include the likes of Stephen King, Neil deGrasse Tyson, and Carl Sagan, among many others.

Edge I found several other tech journalists mentioned in this feature as well, including the former edge Editors Casey Newton and Joanna Stern, ex edge Writer Monica Chen, Wiredlauren good, BloombergMark Gorman and Jason Schreier New York Times’ kashmir hill, AtlanticCaitlin Tiffany, Computer gameswes fenlon, GizmodoRaymond Wong, Digital Foundry Founder Richard Leadbetter, Tom’s guide Editor-in-Chief Mark Spoonauer, formerly Rock paper gun Editor-in-Chief Katherine Castle, formerly IGN News Director Kat Bailey. The descriptions of some experts contain inaccurate information, such as outdated job titles, which could have been accurately updated had Superhuman asked those people for permission to reference their work.

In a statement to EdgeAlex Gay, Vice President of Product and Enterprise Marketing at Grammarly Parent company SuperhumanHe commented, “Expert Review does not seek endorsement or direct involvement from these experts; it offers suggestions inspired by the work of experts and points users toward influential voices who can then explore their knowledge more deeply.”

When asked if Superhuman would consider notifying people mentioned in its AI feature, or asking for their permission, Jay said: “Experts appear in Expert Review because their published work is publicly available and widely cited.”

However, the work of experts has proven difficult to “explore more deeply.” The feature crashes frequently and its “sources” link to spam copies from legitimate websites, or other archived copies that are not the actual source page.

Some sources even went to completely unrelated links that were not written by the person whose work they were supposed to be an example of, which may indicate that the suggestions Grammarly’s AI is making in one person’s name may be based on someone else’s work. This only appears if users click “See More” to expand the suggestions, and then click the “Source” button at the end of the suggestion.

In addition, the way suggestions are presented may be misleading. In Google Docs, the suggestions look similar to comments from real users, and appear to mimic the experience of receiving edits from any expert the AI ​​imitates. One suggestion from Grammarly’s AI is “Inspired by” edge Senior Editor Sean Hollister was all about adding parentheses with context that was already included elsewhere. The only problem is that I was actually edited by the real Sean Hollister, who prefers to avoid redundant or unnecessary explanations while using straightforward wording and organization.

If I had taken this advice and implemented it, the real Sean would probably have removed the parentheses suggested by Grammarly. An AI might be able to ingest massive amounts of someone’s writing and learn to imitate it, sure, but the same strategy can’t teach an AI how to edit the way that person does, based solely on the writing they’ve published, even if you give the bot a check mark on the logo and call it an “expert.”