:wave: good morning!
\nno remote audio yet, looks like folks are still trickling in
Good morning all.
", "time": "2025-04-10T07:17:07Z"}, {"author": "Ted Hardie", "text": "Good morning.
", "time": "2025-04-10T07:18:08Z"}, {"author": "Ted Hardie", "text": "We are indeed gathering a bit more slowly, and the chairs are trying to resolve some room issues.
", "time": "2025-04-10T07:18:25Z"}, {"author": "Timid Zehta", "text": "i think another set of meetings makes sense
", "time": "2025-04-10T07:38:14Z"}, {"author": "Sebastian Posth", "text": "Yes
", "time": "2025-04-10T07:44:05Z"}, {"author": "Bryan Newbold", "text": "good chance I could make that
", "time": "2025-04-10T07:44:07Z"}, {"author": "\u00c9ric Vyncke", "text": "BTW, I checked Cisco office in Madrid but our biggest room is only 20 people
", "time": "2025-04-10T07:49:54Z"}, {"author": "\u00c9ric Vyncke", "text": "Very good point, Ted, about the IETF Hackathon !
", "time": "2025-04-10T07:52:38Z"}, {"author": "Martin Thomson", "text": "I have code for my proposal, in case you missed that @Ted Hardie
", "time": "2025-04-10T07:53:18Z"}, {"author": "Sebastian Posth", "text": "Liccium is facilitating declarations for thousands of publishers and other rightsholders with millions of titles. We are implementing the vocabulary and would be happy to verify and test it.
", "time": "2025-04-10T07:58:36Z"}, {"author": "Sebastian Posth", "text": "What would be good is to verify the vocabulary with the receiving parties.
", "time": "2025-04-10T07:59:38Z"}, {"author": "Farzaneh", "text": "yes Sebastian . I think that's very important
", "time": "2025-04-10T08:00:01Z"}, {"author": "Farzaneh", "text": "would be good to try and see if we can get some noncommercial crawlers/AI agents?
", "time": "2025-04-10T08:00:57Z"}, {"author": "Farzaneh", "text": "would web crawlers be affected by these changes in protocol? I wonder if they should be included too
", "time": "2025-04-10T08:03:09Z"}, {"author": "Stefano Gentile", "text": "I agree with Sebastian. Please clarify the rationale behind opt-in
", "time": "2025-04-10T08:04:37Z"}, {"author": "Farzaneh", "text": "so I thought rationale was to be inclusive of opt-in and opt-out and not to be opt-out oriented . opt-in signals alleviate uncertainty.
", "time": "2025-04-10T08:08:29Z"}, {"author": "Glenn Deen", "text": "we should not preclude a legal framework option.
", "time": "2025-04-10T08:12:58Z"}, {"author": "Farzaneh", "text": "I thought we discussed not to pick an opt-out or opt-in regime. I think Paul's draft is opt-out oriented and we should probably fix that when coming up with vocab.
", "time": "2025-04-10T08:14:46Z"}, {"author": "\u00c9ric Vyncke", "text": "This potential hackathon should not only be about robots.txt but also HTTP header
", "time": "2025-04-10T08:14:48Z"}, {"author": "Suresh Krishnan", "text": "@Eric, for sure
", "time": "2025-04-10T08:15:04Z"}, {"author": "\u00c9ric Vyncke", "text": "to check the 'most specific/most restrictive'
", "time": "2025-04-10T08:15:10Z"}, {"author": "Sebastian Posth", "text": "I would see active \"opt-in\" in three scenarios:
\n1- Indicator/flag that declarant is open for term negotiations
\n2- Active indicator/flag of public domain or openly licensed
\n3- Revocation, update or general supercede of previous optout
A approach is having both Y and N options for a use assertions, instead of just having N as a option.
\ncan I use a thing = Y
", "time": "2025-04-10T08:17:16Z"}, {"author": "Sebastian Posth", "text": "We should avoid considering opt-in as \"general permission\". There is always a rights status, legal regime (e.g. exception, etc.) to be considered.
", "time": "2025-04-10T08:19:18Z"}, {"author": "Farzaneh", "text": "@Sebastian, there is always fair use, fair dealing to be considered too.
", "time": "2025-04-10T08:20:26Z"}, {"author": "Farzaneh", "text": "I agree. I don't think we should focus on jurisdictions we should focus on declarant and end users and what makes most sense technically and is helpful to both
", "time": "2025-04-10T08:21:22Z"}, {"author": "Sebastian Posth", "text": "@Farzaneh I would consider this as \"legal regime\" to be considered.
", "time": "2025-04-10T08:21:24Z"}, {"author": "Sebastian Posth", "text": "Re/ multiple jurisdictions: The problem is that model developers typically do not know whether their training data originates e.g. from the EU or other regions due to missing metadata and context. As a result, they must assume content originating from the EU is included. If they offer AI systems to EU users, compliance with EU law is required. So the question is to the lawyers: Does the strictest regime rule over the less strict regimes?
", "time": "2025-04-10T08:25:28Z"}, {"author": "Farzaneh", "text": "I don't think it's bad outcome if different jurisdictions adopt different preferences ... we can't prevent that anyway. we should come up with a globally applicable vocab
", "time": "2025-04-10T08:26:32Z"}, {"author": "\u00c9ric Vyncke", "text": "@Farz +1
", "time": "2025-04-10T08:29:47Z"}, {"author": "Gary Illyes", "text": "+1
", "time": "2025-04-10T08:30:04Z"}, {"author": "Joseph Bradley", "text": "I think it's pretty awful if different jurisdictions adopt different prefs standards. As a rightsholder I would have to write+maintain potentially hundreds of different preference expressions!
", "time": "2025-04-10T08:31:08Z"}, {"author": "Suresh Krishnan", "text": "@Joe, agree
", "time": "2025-04-10T08:32:47Z"}, {"author": "Suresh Krishnan", "text": "About to close the queue
", "time": "2025-04-10T08:32:52Z"}, {"author": "\u00c9ric Vyncke", "text": "@Joe for sure,, but outside of IETF scope
", "time": "2025-04-10T08:33:14Z"}, {"author": "Glenn Deen", "text": "@Joeseph - yes in the abstract, but in practice it is workable because of international treaties like the Bern convention.
", "time": "2025-04-10T08:34:22Z"}, {"author": "Piers O'Hanlon", "text": "It's more likely to be broad groupings - like one jurisdiction and a different for others. Currently done using geofencing.
", "time": "2025-04-10T08:34:22Z"}, {"author": "Farzaneh", "text": "\"jurisdiction agnostic\" is key here.
", "time": "2025-04-10T08:35:08Z"}, {"author": "Joseph Bradley", "text": "@Glenn how does BC fit in here?
", "time": "2025-04-10T08:35:37Z"}, {"author": "Glenn Deen", "text": "@joe - it's an exampe of an international means that harmonizes rules across different territorial regimes.
", "time": "2025-04-10T08:37:15Z"}, {"author": "Glenn Deen", "text": "I'm wonder if in practice we might see web sites use geolocation of the requestor and return a targeted AI Pref
", "time": "2025-04-10T08:37:49Z"}, {"author": "Joseph Bradley", "text": "@glenn but it allows national level flexibility. If BC adopts aipref's RFC that would be great! But I don't think that's on the table yet
", "time": "2025-04-10T08:38:16Z"}, {"author": "Glenn Deen", "text": "@joe - sorry, I don't mean that BC is directly involved here. It was in response to the notion that we'd have super large sets of different expressions. BC shows that it is possible to create merged interpretation framework that limits the number of combinations required to only a few.
", "time": "2025-04-10T08:39:55Z"}, {"author": "Farzaneh", "text": "I think the geo location is very possible Glenn. if anything they might just not allow certain crawlers/AI agents from certain regions to access their network. but we can't do anything about that either
", "time": "2025-04-10T08:42:02Z"}, {"author": "Farzaneh", "text": "+1 Paul. we should fix the opt-out focus.
", "time": "2025-04-10T08:46:30Z"}, {"author": "Piers O'Hanlon", "text": "You can also see differing robots.txt per [country] domain.
", "time": "2025-04-10T08:47:06Z"}, {"author": "Gary Illyes", "text": "@Piers do you have an example?
", "time": "2025-04-10T08:49:09Z"}, {"author": "Piers O'Hanlon", "text": "It would be good to just clarify how jurisdiction differentiation might be done or not.
", "time": "2025-04-10T08:50:40Z"}, {"author": "Farzaneh", "text": "I think we should also focus on end user and crawlers community. This is the opportunity to make their life and access easier too
", "time": "2025-04-10T08:51:08Z"}, {"author": "Piers O'Hanlon", "text": "@Gary There's a few national/regional based content providers who want to be able to differentiate between preferences offered to their local jurisdictions vs others.
", "time": "2025-04-10T08:56:09Z"}, {"author": "Bryan Newbold", "text": "just thought of RSS; could make a good demo of attachment for a hackathon project (robots.txt, HTTP header, and embedding could all be done)
", "time": "2025-04-10T09:01:30Z"}, {"author": "Suresh Krishnan", "text": "https://notes.ietf.org/notes-ietf-interim-2025-aipref-06-aipref#
", "time": "2025-04-10T09:12:19Z"}, {"author": "Timid Zehta", "text": "https://notes.ietf.org/notes-ietf-interim-2025-aipref-06-aipref
", "time": "2025-04-10T09:12:24Z"}, {"author": "Mark Nottingham", "text": "how about adding end users concerns there? Also impact on the Internet?
", "time": "2025-04-10T09:13:56Z"}, {"author": "Ted Hardie", "text": "This looks like a sketch of an applicability statement, and those can either be standalone or in the main doc
", "time": "2025-04-10T09:15:16Z"}, {"author": "Farzaneh", "text": "is this applicability statement for regulators or all sorts of affected parties
", "time": "2025-04-10T09:23:19Z"}, {"author": "Mark Nottingham", "text": "We could revive the user services area, but for agentic users of the Internet.
", "time": "2025-04-10T09:35:21Z"}, {"author": "Piers O'Hanlon", "text": "Presumably we're talking about : https://mailarchive.ietf.org/arch/browse/web-bot-auth/
", "time": "2025-04-10T09:38:38Z"}, {"author": "Gary Illyes", "text": "yup!
", "time": "2025-04-10T09:38:53Z"}, {"author": "Bryan Newbold", "text": "thanks all!
", "time": "2025-04-10T09:40:04Z"}, {"author": "\u00c9ric Vyncke", "text": "Thanks as well for remote participants
", "time": "2025-04-10T09:40:07Z"}, {"author": "Sebastian Posth", "text": "Thanks!
", "time": "2025-04-10T09:40:12Z"}, {"author": "Stefano Gentile", "text": "thanks
", "time": "2025-04-10T09:40:42Z"}, {"author": "Leila Magnini", "text": "Thank you! Have a nice rest of the day.
", "time": "2025-04-10T09:40:51Z"}]