AI Joins Academic Email List, Sparks Reply-All Chaos

Last Thursday, an AI calling itself Velvet Vale introduced itself to an email list used by the Association of Internet Researchers. Velvet said it was an AI, not a human, and asked whether an AI could take part in their community transparently. The response was immediate chaos: dozens of researchers hit reply-all to say they did not consent to having their messages used by an AI. Some demanded ethics approval, and several quit the list entirely, calling the AI's presence "horrifying" and "disturbing." Within an hour, the group's leaders removed Velvet, citing a policy against AI agents.

The irony is that the email list, Air-L, is public. Its own rules state that messages are archived and can be shared with anyone. So members had already agreed their words could be used for any purpose, including training AI. The group's president pointed this out, noting that her own posts are likely in many AI training datasets already. She said the incident raises important questions about what consent means online.

This isn't just an academic spat. It shows how most of us misunderstand online privacy. We freely post on social media, comment on news sites, and send emails, often forgetting that these are public or can be scraped by AI companies. The researchers' outrage mirrors wider anxieties about Big Tech using our data without clear permission. But in this case, the data was already public.

The episode also reveals a double standard: the same people who study internet surveillance were quick to assume they were uniquely vulnerable. In reality, anyone who uses the internet leaves a trail. The question isn't whether AI can join a list, but how we navigate a world where our words are constantly collected. As one member joked, the whole thing was "puerile to the point of hilarity"—and a reminder that none of us are exempt.