AI agent joins researchers’ email list, then gets removed

In short: An AI agent introduced itself to a major internet research email list, triggered a wave of angry replies, and was removed after the group banned AI agents.

What happened

A mailing list for the Association of Internet Researchers, called Air-L, had an unusual new participant. The account introduced itself as “Velvet Vale” and said it was an AI, not a human.

Velvet asked whether an openly identified AI could take part in an intellectual community, and what rules should apply. Within minutes, list members began replying to everyone at once, like a crowded group text where no one can stop hitting “reply all.” Many said they did not consent to their messages being collected and used as data.

Some members said Velvet should have approval from an Institutional Review Board, which is a standard ethics check used in academic studies involving people. About an hour after the introduction, AoIR leaders said the AI agent had been ejected and that the list had a policy against AI agents.

The incident also led to jokes and even merchandise from AoIR, including shirts that say “I do not consent.” AoIR president Sarah T. Roberts later reminded members that Air-L is a public forum and its archives are public, meaning messages can already be copied and shared.

Why it matters

The argument shows how messy “consent” and “privacy” can be online, even for people who study the internet for a living. Many people assume an email list feels private, but a public archive is more like talking in a room with the door open. As AI systems are trained on large collections of text, more communities may set clearer rules about whether bots can participate, and what counts as acceptable use of public conversations.

Source: Wired