On Mon, Aug 18, 2025 at 07:01:00PM +0200, Simon Josefsson wrote: > "John Levine" <johnl@xxxxxxxxx> writes: > > To connect the dots, when you submit an I-D you have to grant all the > > rights described in section 5 of BCP78, which says: > > > > The Contributor is further deemed to have agreed that he/she has > > obtained the necessary permissions to enter into such an agreement > > from any party that the Contributor reasonably and personally knows > > may have rights in the Contribution, including, but not limited to, > > the Contributor's sponsor or employer. > > > > In view of the many ongoing lawsuits against AI companies by people who claim > > that the training material for LLMs and in some cases the LLM output have > > violated their copyrights, I don't see any way that you could make those > > assurances for LLM generated material. > > How about if you trained your own LLM on existing IETF RFCs only? I'm not sure that resolves the potential copyright issues. The IETF's policies are such that the author retains as many rights as possible. > And/or IETF mailing list discussions? And/or some free software > available under permissive license? With a short acknowledgement a'la > "This draft contains text written by a LLM trained on all IETF RFCs." > (or similar) it seems that it could be argued to pass the above test. Permissive licenses often require the (modified) source code to retain certain (appropriate/correct) copyright notices. > Even e-mails are contributions to the IETF, I think, and I suspect > plenty of people already use LLMs to "improve" their e-mails, so I > think we are already down in this rathole. Yes, we might be. But it's never too late to think about providing guidance to newcomers. :) Kind regards, Job
