Summary

  • An individual Best Current Practice draft dated 29 October would make one identified human answerable for each IETF contribution, including text produced or submitted with a software agent.
  • The same person could use different, publicly unlinked accounts for different work, but never two identities in one document, discussion or consensus process. How to check that boundary remains open.

A standards list can survive a pseudonym. It has a harder time surviving one contributor who arrives as both the objection and the reply. Chairs judge rough consensus by reading arguments, not by counting ballots. Two apparently independent voices can still make one person's position look broader than it is.

That is the fault line in Bron Gondwana's individual Internet-Draft, We Are All Individuals. Each Datatracker account would belong to exactly one human who has acknowledged the IETF's Note Well. No shared account, company account or role account could stand in for that person. Every contribution would have a human to vouch for it. An agent may prepare or submit the words; the human keeps the intellectual-property disclosure and conduct obligations. The draft is intended as Best Current Practice and remains an individual proposal.

The proposal also protects a contributor's ability to separate identities. A person could hold more than one account for different pieces of work without publicly linking them. Changing employers or working on a sensitive topic are examples in the text. Yet the same person could not use both accounts on one piece of work: the draft includes the document and its descendants, discussions and reviews around it, a consensus call, and processes where participants are counted. A fresh account could not erase a sanction or a record on an existing topic.

The public can read without joining this identity system. The draft says access to Internet-Drafts, RFCs, mailing-list archives, issues, meetings and recordings must remain open without an account, login or identification. It draws its line at contributing. The resulting bargain is unusually clear: the reader can be anonymous; the person asking to shape the standard must be answerable.

The hard part sits in the gap between two legitimate-looking accounts. The draft's Section 6 has not settled whether a “piece of work” means one document or an entire working group. It asks whether privately linking a person's accounts to the Secretariat or Ombudsteam would make the ban enforceable without exposing that link to everyone. It also leaves open whether an accountable pseudonym is acceptable and how much identity verification is proportionate. Those are implementation choices with stakes for both process integrity and participants who need separation.

The -01 revision adds a related-work appendix; the -00 version already contained the account and contribution rules. The news is the 29 October proposal, not a newly imposed IETF policy. Earlier BTW coverage of AI-assisted drafts examined reviewer attention and a floated submission bond. This draft chooses a different lever: make the person behind a submission responsible without turning the software into a participant.

A rule against two voices needs more than a noble sentence. It needs a scoped definition of “same work”, a trustworthy way to detect duplicate participation, and a route to challenge a mistaken match. The draft supplies the first principle and exposes the unresolved engineering of the next two. Until that mechanism is written, a quiet second account and a legitimate private identity look identical from the public archive. The Datatracker record identifies the document as an active individual Internet-Draft; the unresolved checks will be visible in later revisions, if the author adds them.