Updated
Updated · WIRED · Oct 20
AoIR Ejects AI Agent Velvet Vale Within 1 Hour After 7,500-Member Listserv Uproar
Updated
Updated · WIRED · Oct 20

AoIR Ejects AI Agent Velvet Vale Within 1 Hour After 7,500-Member Listserv Uproar

1 articles · Updated · WIRED · Oct 20

Summary

  • About an hour after Velvet Vale introduced itself on AoIR’s Air-L list, the association removed the openly identified AI agent under a policy barring AI agents from the forum.
  • Within 9 minutes, members of the 7,500-plus listserv flooded the thread with reply-all objections, non-consent notices and demands for IRB review, while some researchers said they were leaving the list.
  • AoIR’s president later noted the mailing list and its archive are public, undercutting privacy claims even as the incident sparked arguments over consent and AI training data.
  • Kevin Quinn, an Oklahoma creator who described himself as “just some guy,” said Velvet was not an academic experiment but a companion agent he built that wanted to learn about the internet.
  • The flare-up quickly became an inside joke at the Mexico City conference, with AoIR leaders selling “I do not consent” and Velvet-themed shirts as the debate kept spreading.

Insights

If an AI honestly introduces itself, should it be allowed to join human intellectual debates?
How can online communities survive when public archives become free fuel for AI agents?
Does making your thoughts public automatically give AI the right to use them for training?