Re: Dealing with LLMs in IETF discussions draft

Hi Laurence,
At 11:09 AM 31-07-2026, Laurence Lundblade wrote:
>My top-of-mind reaction: people are responsible for what they post.

[snip]

>If you blast out lots of AI-originated text that you haven't 
>reviewed thoroughly and maybe don't even understand, your 
>credibility in the IETF is going to decline. At some level, not 
>understanding what you post even becomes a basis for moderators 
>prohibiting posts from you.

The approach over the years was that there were people sending the 
messages to the various mailing list to propose, comment, or review 
the submissions.  There were different barriers which a few 
participants faced, e.g. the participant had to use the English 
language to communicate, the participant had to do a lot of 
reading/research to reach a level which other participants viewed as 
acceptable.  A participant who was unable to satisfy the criteria 
would be met with ridicule.  A participant who was well-connected 
would get a pass.

The scenario which you wrote about might be a clear-cut case.  There 
are some participants who might be using Large Language Models to 
avoid those potential issues.  There are some participants who would 
be using Large Language Models as it's a good shortcut to getting 
their drafts published.  It's difficult to figure out whether a 
person is in the first group or the second group when that person 
doesn't have an existing online footprint.  Moderating someone in the 
first group could be viewed as prejudiced.  Not moderating the person 
means going against the subscribers who sent private complaints.

Regards,
S. Moonesamy