14:58:05 <RRSAgent> RRSAgent has joined #webmachinelearning
14:58:09 <RRSAgent> logging to https://www.w3.org/2026/04/02-webmachinelearning-irc
14:58:09 <Zakim> RRSAgent, make logs Public
14:58:10 <Zakim> please title this meeting ("meeting: ..."), anssik
14:58:10 <anssik> Meeting: WebML CG Teleconference – 2 April 2026
14:58:14 <anssik> Chair: Anssi
14:58:18 <anssik> Agenda: https://github.com/webmachinelearning/meetings/blob/main/telcons/2026-04-02-cg-agenda.md
14:58:18 <lgombos> lgombos has joined #webmachinelearning
14:58:21 <anssik> Scribe: Anssi
14:58:27 <anssik> ScribeNick: anssik
14:59:15 <anssik> Present+ Anssi_Kostiainen
14:59:22 <anssik> Regrets+ Dominic_Farolino
14:59:39 <anssik> Regrets+ Christian_Liebel
14:59:39 <kush> kush has joined #webmachinelearning
15:00:00 <BenGreenstein> BenGreenstein has joined #webmachinelearning
15:00:06 <anssik> Present+ Khushal_Sagar
15:00:08 <jaeone> jaeone has joined #webmachinelearning
15:00:21 <anssik> Present+ Winston_Chen
15:00:26 <winstonc> winstonc has joined #webmachinelearning
15:00:26 <anssik> Present+ Laszlo_Gombos
15:00:35 <anssik> Present+ Jaewon_Lee
15:00:45 <anssik> Present+ Benjamin_VanderSloot
15:01:03 <lgombos> Present+ Laszlo_Gombos
15:01:08 <anssik> Present+ Mark_Foltz
15:01:13 <anssik> Present+ Ben_Greenstein
15:01:39 <winstonc> Present + Winston_Chen
15:02:07 <Mark_Foltz> Mark_Foltz has joined #webmachinelearning
15:02:18 <anssik> Anssi: welcome to the new participants to the group:
15:02:22 <anssik> RRSAgent, draft minutes
15:02:23 <RRSAgent> I have made the request to generate https://www.w3.org/2026/04/02-webmachinelearning-minutes.html anssik
15:02:36 <anssik> ... Joshua Lochner and Tarek Ziade from Hugging Face
15:02:45 <johannhof> johannhof has joined #webmachinelearning
15:02:51 <anssik> ... Yoav Weiss from Shopify
15:02:56 <anssik> ... Rupert Manfredi from Telepath
15:03:01 <anssik> ... Chelsea Kohl from Flying Okole
15:03:10 <anssik> ... Joan Leon and David Mulder as individual contributors
15:03:14 <anssik> ... welcome all to the WebML Community Group!
15:03:21 <anssik> ... as a reminder, we'll use IRC-based queue management in this meeting:
15:03:26 <anssik> -> https://www.w3.org/guide/meetings/zakim.html#speakerqueue
15:03:29 <anssik> q+
15:03:31 <anssik> ack anssik
15:03:36 <anssik> Anssi: to suggest agenda topics, use Agenda+ label
15:03:47 <anssik> -> Agenda+ https://github.com/webmachinelearning/webmcp/labels/Agenda+
15:03:51 <saron> saron has joined #webmachinelearning
15:04:12 <anssik> Topic: Announcement: Workshop on E-commerce for Humans and AI Agent
15:04:12 <iris> iris has joined #webmachinelearning
15:04:15 <anssik> Anssi: a new "Workshop on E-commerce for Humans and AI Agent" in preparation, goals:
15:04:27 <anssik> ... "This workshop aims to share experience of creating content with AI Agents in mind."
15:04:31 <anssik> ... "The impetus for the workshop is from ecommerce which will be a particular focus of the event."
15:04:36 <anssik> ... I contributed the WebMCP shopping use case for consideration:
15:04:40 <anssik> -> https://github.com/webmachinelearning/webmcp#example---shopping
15:04:46 <benvds> benvds has joined #webmachinelearning
15:04:47 <anssik> Anssi: for more information, and general questions and comments, see the GH issue:
15:04:52 <anssik> -> https://github.com/w3c/strategy/issues/544
15:04:52 <gb> https://github.com/w3c/strategy/issues/544 -> Issue 544 Workshop: ECommerce for humans and AI Agents (joint workshop with GS1) (by dontcallmedom)
15:05:00 <victor> victor has joined #webmachinelearning
15:05:01 <anssik> Anssi: I will share more information when this workshop is officially announced
15:05:15 <anssik> Topic: Writing Assistance APIs
15:05:19 <anssik> gb, this is webmachinelearning/writing-assistance-apis
15:05:20 <gb> anssik, OK.
15:05:33 <anssik> Subtopic: New parameter to support model tiering in Summarizer
15:05:37 <anssik> Anssi: issue #96
15:05:38 <gb> https://github.com/webmachinelearning/writing-assistance-apis/issues/96 -> Issue 96 New parameter to support model tiering in Summarizer (by jingyun19) [Agenda+]
15:05:46 <anssik> ... Jaewon Lee from Google will present the proposal
15:05:50 <anssik> ... he has produced an explainer document that contains the latest proposal:
15:05:54 <anssik> -> https://github.com/webmachinelearning/writing-assistance-apis/blob/main/explainers/summarizer-preference-param.md
15:06:01 <kush> kush has joined #webmachinelearning
15:06:24 <anssik> Jaewon: I work on Web AI team at Google, working with Reilly and Mike
15:06:45 <anssik> ... presenting model tiering proposal that can apply to multiple APIs, initially summarizer
15:07:03 <anssik> ... currently, developers have no way to signal to the browser that they are willing to trade off some output quality
15:07:05 <anssik> ... or specific capabilities for faster execution or broader device compatibility
15:07:17 <anssik> ... proposing a new parameter, preference, to the SummarizerCreateCoreOptions
15:07:23 <anssik> [[
15:07:23 <anssik> enum PerformancePreference {
15:07:23 <anssik>   "auto",
15:07:23 <anssik>   "speed",
15:07:23 <anssik>   "capability"
15:07:23 <anssik> };
15:07:23 <anssik> ]]
15:07:37 <jason> jason has joined #webmachinelearning
15:07:49 <brwalder> brwalder has joined #webmachinelearning
15:08:20 <anssik> Jaewon: "capability" for the most capable mode, "speed" for the fastest response, "auto" for the implementation to decide
15:08:59 <anssik> ... three things, this helps with a11y and reach, UX with faster interaction, graceful degradation with availability check
15:09:13 <anssik> Anssi: thank you Jaewon for the introduction to this proposal
15:09:27 <anssik> Anssi: is the group supportive of advancing this proposal to the spec PR?
15:09:29 <anssik> q?
15:09:48 <Ehsan> Ehsan has joined #webmachinelearning
15:09:55 <anssik> Anssi: how about implementation exprience at this stage?
15:10:17 <kush> q+
15:10:17 <anssik> Jaewon: we are working on a smaller model currently, not yet tested this with a bigger model
15:10:17 <anssik> q?
15:10:19 <anssik> ack kush
15:10:36 <anssik> Kushal: do developers get feedback on what model will be selected?
15:11:13 <anssik> Jaewon: we don't currently expose the model
15:11:15 <anssik> q?
15:11:16 <Mark_Foltz> q+
15:11:22 <anssik> ack Mark_Foltz
15:11:45 <anssik> Mark: I wondered if the idea was to set preference to speed to make sure you always get the fastest model?
15:12:18 <anssik> Jaewon: if you choose speed now, in Chrome you'd have two models to choose from, in the future we might have a larger variety of models
15:12:40 <anssik> Mark: I'm wondering if in 3 model scenario you'd always have a speedy model option
15:13:08 <anssik> Jaewon: the logic is currently implemention-defined
15:13:24 <anssik> q?
15:14:32 <Mark_Foltz> +1
15:14:40 <iris> +1
15:14:44 <lgombos> +1
15:14:47 <anssik> RESOLUTION: Add a three-tier performance preference feature to the Summarizer API. (issue #96)
15:15:08 <anssik> Topic: Translator and Language Detector APIs
15:15:13 <anssik> gb, this is webmachinelearning/translation-api
15:15:13 <gb> anssik, OK.
15:15:18 <anssik> Subtopic: Definition of "On-Device Models"
15:15:42 <anssik> Anssi: Ehsan shared a draft explainer that attempts to define "On-Device Models"
15:15:47 <anssik> -> https://github.com/SamsungInternet/Explainers/blob/main/drafts/demense-control.md
15:16:20 <anssik> Anssi: problem statement, inference may run locally or on remote servers
15:16:29 <anssik> ... users have varying privacy/security needs wrt this
15:16:42 <anssik> ... Web APIs have no standard way to express or enforce such preferences
15:17:09 <anssik> ... proposed solution: API extension to allow the User Agent to constrain where inference runs based on user preference, two preference knobs proposed:
15:17:15 <anssik> ... 1) execution locations:
15:17:23 <anssik> ... personal-secure = "personal device" + "secure area"
15:17:34 <anssik> ... personal = "personal device"
15:17:47 <anssik> ... remote-secure = "remote device" + "secure area"
15:17:52 <anssik> ... remote-attested - "remote device" + "legal/contractual guarantees"
15:18:03 <anssik> ... remote - "remote device" + "no guarantees"
15:18:14 <anssik> ... 2) direction preference:
15:18:29 <anssik> ... more-privacy-preserving = allow upgrade to more privacy-preserving location
15:18:34 <anssik> ... less-privacy-preserving = allow fallback to less privacy-preserving location
15:18:42 <anssik> ... key design principles:
15:18:52 <anssik> ... - user preferences take precedence over developer preferences
15:19:03 <anssik> ... - UA picks the most privacy-preserving location both parties accept
15:19:19 <anssik> ... - fingerprinting risk mitigated via user activation or rate limiting
15:19:36 <anssik> q?
15:20:19 <anssik> Ehsan: Ehsan from Samsung, this is a follow-up to an email I sent to the group earlier, TAG response to Translation API and its network usage
15:20:40 <anssik> ... that led to a discussion what we mean by "on-device models" actually, and explainer drafting followed
15:20:53 <anssik> ... TAG wanted to hear this groups opinion on this direction
15:21:19 <anssik> ... TAG sees a lot of features using on-device models so this would generalize
15:21:20 <anssik> q?
15:21:50 <anssik> q?
15:22:25 <anssik> Anssi: given this is not a testable feature, it cannot be a normative requirement
15:22:45 <anssik> ... I wonder if this mechanism has value if it is informative hints?
15:23:09 <AlexN> AlexN has joined #webmachinelearning
15:23:41 <anssik> Ehsan: my initial response is there's a compliance framework, if compliance for service provider is explained we can infer indirectly about it
15:23:49 <Mark_Foltz> q+
15:23:54 <anssik> ack Mark_Foltz
15:24:36 <anssik> Mark: I think potentially as a user I'd like to know if there's a way to expose from a platform I trust what compute unit is actually used
15:24:58 <anssik> ... that might be out of scope of the API spec but worth discussing in the explainer
15:25:35 <anssik> ... some kind of audit log might be a solution
15:26:04 <anssik> Anssi: this feature is similar to the DNT (Do Not Track) feature that is no longer recommended due to challenges in enforcing websites to comply
15:26:59 <anssik> Ehsan: DNT came up in TAG discussion and we are trying to learn from those mistakes
15:27:11 <anssik> ... thank you for resurfacing that issue
15:28:08 <anssik> Ehsan: the issue is open for further feedback, I hear from this group the general sense is we are in the right direction but not quite ready with the exact mechanism or API shape
15:28:44 <anssik> Anssi: maybe use case focus would help
15:28:54 <anssik> Topic: WebMCP
15:28:58 <anssik> gb, this is webmachinelearning/webmcp
15:28:59 <gb> anssik, OK.
15:29:06 <anssik> Subtopic: Allow iframes to declare tools
15:29:08 <anssik> Anssi: issue #57
15:29:09 <gb> https://github.com/webmachinelearning/webmcp/issues/57 -> Issue 57 Allow iframes to declare tools (by khushalsagar) [Agenda+]
15:29:53 <anssik> Anssi: in our last meeting we resolved to explore Document Policy as a bi-directional opt-in mechanism without a wildcard mechanism that addresses the two use cases, figure out a solution for name collisions
15:29:58 <anssik> -> https://www.w3.org/2026/03/19-webmachinelearning-minutes.html#ba6d
15:30:07 <anssik> Anssi: as a reminder, we target two use cases:
15:30:21 <anssik> ... 1. To allow iframes to declare tools for browser/extension agents to operate
15:30:28 <anssik> ... 2. To allow some form of cross frame calling of tools
15:30:34 <anssik> ... since the resolution was made, new information has been shared:
15:30:55 <anssik> ... Dominic re-raised MarkF's comment for the need to come up with a serialization scheme to disambiguate name collisions across frames
15:31:08 <anssik> ... Alex noted we can use postMessage to relay tools between host and child to address use case 2
15:31:36 <anssik> Anssi: Victor shared a concrete example for use case 1, embedded Calendry automated scheduling tool that does not work with postMessage only:
15:31:56 <anssik> ... a user adds <iframe src="https://calendly.com/your-link"> to their site and expects the WebMCP tools provided by Calendry to be seen by the UA to be able book a meeting
15:32:04 <anssik> ... this should not require setting up a postMessage bridge, <iframe> embed should be enough
15:32:15 <anssik> ... but would still require Calendry to explicitly opt in to exposing tools
15:32:31 <anssik> ... Johann notes security benefits of embedding iframe vs embedding JS
15:32:36 <simone> present+
15:32:42 <anssik> ... Victor asked Ian for insights re Document Policy vs Permission Policy, no response yet
15:33:07 <AlexN> +1
15:33:15 <anssik> ... does the group agree the Calendry use case (concrete example of use case 1) is important to support?
15:33:52 <anssik> Victor: Calendry has been the key use case for this feature, a lot of people prefer to embed iframes, copy-paste skippets to websites
15:34:21 <anssik> ... it'll be great if tools like this could expose WebMCP tools, the parent does not care about these tools to be exposed
15:34:46 <kush> q+
15:35:20 <anssik> ... one thing what came out, consider a case where the service says "I allow anyone that embeds me to access my tools"
15:35:21 <anssik> q?
15:35:50 <anssik> Anssi: do we want to await for further data on Document Policy vs. Permission Policy?
15:36:31 <anssik> Victor: may concern is that firstly, in general Document Policy is the Right Thing, more ergonomic
15:36:43 <Mark_Foltz> q+
15:36:51 <anssik> ... and on the embeddee site also "I allow this site to embed me"
15:37:13 <anssik> ... Document Policy is only implemented in Chromium, the questions is should take a dependency on that
15:37:38 <anssik> ... alternative is to use a mix of Permission Policy and bake into the WebMCP spec "I allow this website to use me" mechanism
15:37:41 <anssik> q?
15:37:53 <anssik> ack kush
15:38:36 <anssik> Kushal: relaying this from Dominic, one question he had, for Mozilla folks, RFP for Document Policy was neutral, are you interested in this feature?
15:39:24 <anssik> Benjamin: part of the reason it is neutral is that there wasn't many applications for that feature, this is now one such application
15:40:04 <anssik> Benjamin: I will follow up on that Document Policy Mozilla position
15:40:06 <anssik> q?
15:40:08 <anssik> ack Mark_Foltz
15:40:31 <kush> q+
15:41:10 <anssik> Mark: comment on the Calendry use case, makes a lot of sense, I perceive that with Document Policy also emberred needs to opt in, in that case for this new capability means Calandry would need to update its code
15:41:25 <anssik> ack kush
15:41:29 <victor> q+
15:41:29 <johannhof> q+
15:41:47 <anssik> Kushal: question about name collision aspect, what's the issue there?
15:41:54 <Mark_Foltz> q+
15:42:11 <anssik> ... we have to deal with docs that have the same name and deal with that as a consuming agent, two docs with the same name, why this needs to be solved on the platform side
15:42:16 <anssik> ack victor
15:42:29 <Mark_Foltz> Site agents may see name collisions from embeds
15:42:57 <anssik> Victor: mostly came from the 2nd use case, website A embeds an iframe and is able to see the tools on another website B
15:43:23 <anssik> ... some sort of origin key needed for interaction between iframe and website in this scenario
15:43:48 <anssik> Kushal: either origin-based namespacing or document-based namespacing is needed
15:44:28 <anssik> Victor: I thought iframe gives parent frame access to all tools
15:44:42 <anssik> Kushal: consumer needs to know the document id and tool name
15:44:44 <anssik> q?
15:45:07 <anssik> Kushal: how do we deal with nested iframes?
15:45:21 <anssik> ack johannhof
15:45:49 <anssik> Johann: Permissions Policy bubbles up, not sure how Document Policy works in that regard
15:46:02 <anssik> ... following up on Mark's comment for widcard
15:46:29 <anssik> ... are we reducing the utility of the API by making it so straight, top-level should not be able to say " I accept tools from any 3rd party"
15:46:59 <anssik> ... on the 3rd party side, I'd be more flexible if we can make it so web developers understanding opting in to share with all is potentially dangerous
15:47:13 <anssik> ... want to avoid the attack case that wildcard allows
15:47:26 <anssik> ... category of attack where you can iframe someone else
15:47:27 <anssik> q?
15:47:30 <anssik> ack Mark_Foltz
15:48:23 <anssik> Mark: about name collision issue, two aspects, one on the browser agent side, multiple frames and multiple browsing contexts, how to organize tool names in namespacing, origin-keys don't address all edge cases
15:48:42 <anssik> ... for the API space this is about site agents, to enumerate tools across document tree
15:48:56 <anssik> ... how to present the enumeration to site agents to disambiguate tools
15:49:17 <anssik> ... ideally, the frame providing the tools could provide a hint so the model understand from where they came from
15:49:18 <anssik> q?
15:50:35 <kush> q+
15:51:06 <anssik> ack kush
15:52:13 <anssik> Victor: use cases coming up, interested in iframe giving parent access
15:52:17 <anssik> q?
15:52:33 <victor> iframe giving the browser access*
15:53:09 <anssik> Subtopic: Threat modeling
15:53:14 <anssik> Anssi: issue #154
15:53:23 <anssik> Anssi: Security Interest Group has produced a guide to help spec authors and implementers document and understand threats facing the web
15:53:27 <gb> https://github.com/webmachinelearning/webmcp/issues/154 -> Issue 154 Threat modeling (by anssiko) [Agenda+] [security]
15:53:28 <anssik> -> https://www.w3.org/TR/threat-modeling-guide/
15:53:28 <simone> q+
15:53:32 <anssik> ... the guide is WIP thus rough edges expected
15:53:39 <anssik> ... I reviewed the doc and provided feedback via GH issues
15:53:51 <anssik> ... "How to Threat Model" is a three step process:
15:53:56 <anssik> ... 1. Paint a picture of the system
15:54:02 <anssik> ... 2. Capture threats
15:54:09 <anssik> ... 3. Describe responses
15:54:15 <anssik> ... for step 1, the expected deliverable is a data flow diagram, UML sequence diagram or such that illustrates the actors and the flows between them
15:54:19 <anssik> ... WebMCP explainer has a diagram we can reuse as a starting point:
15:54:23 <anssik> -> https://github.com/webmachinelearning/webmcp/blob/main/content/explainer_webmcp.png
15:54:47 <anssik> Anssi: for step 2, I've asked the authors of this guide for enhancements to the guide to help us get started with WebMCP threat identification and analysis
15:54:55 <anssik> ... I think step 3 follows more naturally when we have identified the threats in step 2
15:54:59 <anssik> ... questions, comments?
15:55:02 <anssik> q?
15:55:28 <brwalder> Credit goes to Google for that diagram. I just converted it to PNG.
15:55:32 <anssik> Simone: thanks for the comments for the guide, Anssi
15:56:00 <anssik> ... I love the diagram in the explainer, I also see Johann and others looking at privacy and security aspects of WebMCP
15:56:33 <anssik> ... we have threat modeling group that does threat modeling together on new proposals
15:57:02 <anssik> ... we also work asynchronously
15:57:24 <anssik> ... we can help this group with this exercise
15:57:25 <anssik> q?
15:58:06 <anssik> ack simone
15:58:41 <anssik> Anssi: are there other web features that have done threat modeling we could learn from?
15:59:21 <anssik> Simone: digital credentials is one early feature that has experimented with this threat modeling process
15:59:36 <anssik> q?
16:02:16 <anssik> Johann: I appreciate the exercise, related features are a bit new space, a lot of depend on probabilistic defences for which we don't have solutions in the standards based, we're working on deterministic features
16:02:22 <anssik> ... historically
16:02:47 <AlexN> AlexN has joined #webmachinelearning
16:02:47 <Ehsan> Ehsan has joined #webmachinelearning
16:02:47 <brwalder> brwalder has joined #webmachinelearning
16:02:47 <kush> kush has joined #webmachinelearning
16:02:47 <victor> victor has joined #webmachinelearning
16:02:47 <benvds> benvds has joined #webmachinelearning
16:02:47 <iris> iris has joined #webmachinelearning
16:02:47 <saron> saron has joined #webmachinelearning
16:02:47 <johannhof> johannhof has joined #webmachinelearning
16:02:47 <winstonc> winstonc has joined #webmachinelearning
16:02:47 <jaeone> jaeone has joined #webmachinelearning
16:02:47 <BenGreenstein> BenGreenstein has joined #webmachinelearning
16:02:47 <lgombos> lgombos has joined #webmachinelearning
16:02:47 <reillyg> reillyg has joined #webmachinelearning
16:03:14 <anssik> Simone: next step would be to work asynchronously on the threat modeling of WebMCP and perhaps set up a meeting to work on the problem space
16:03:37 <anssik> s/space/space in realtime
16:04:01 <AlexN> AlexN has joined #webmachinelearning
16:04:01 <Ehsan> Ehsan has joined #webmachinelearning
16:04:01 <brwalder> brwalder has joined #webmachinelearning
16:04:01 <kush> kush has joined #webmachinelearning
16:04:01 <victor> victor has joined #webmachinelearning
16:04:01 <benvds> benvds has joined #webmachinelearning
16:04:01 <iris> iris has joined #webmachinelearning
16:04:01 <saron> saron has joined #webmachinelearning
16:04:01 <johannhof> johannhof has joined #webmachinelearning
16:04:01 <winstonc> winstonc has joined #webmachinelearning
16:04:01 <jaeone> jaeone has joined #webmachinelearning
16:04:01 <BenGreenstein> BenGreenstein has joined #webmachinelearning
16:04:01 <lgombos> lgombos has joined #webmachinelearning
16:04:01 <reillyg> reillyg has joined #webmachinelearning
16:04:18 <anssik> Johann: everyone's very busy working with WebMCP so we appreciate support from Security IG participants with this exercise
16:04:19 <anssik> q?
16:04:59 <anssik> Johann: let's sync in the issue for the exact next steps
16:05:01 <anssik> q?
16:05:24 <AlexN> AlexN has joined #webmachinelearning
16:05:24 <Ehsan> Ehsan has joined #webmachinelearning
16:05:24 <brwalder> brwalder has joined #webmachinelearning
16:05:24 <kush> kush has joined #webmachinelearning
16:05:24 <victor> victor has joined #webmachinelearning
16:05:24 <benvds> benvds has joined #webmachinelearning
16:05:24 <iris> iris has joined #webmachinelearning
16:05:24 <saron> saron has joined #webmachinelearning
16:05:24 <johannhof> johannhof has joined #webmachinelearning
16:05:24 <winstonc> winstonc has joined #webmachinelearning
16:05:24 <jaeone> jaeone has joined #webmachinelearning
16:05:24 <BenGreenstein> BenGreenstein has joined #webmachinelearning
16:05:24 <lgombos> lgombos has joined #webmachinelearning
16:05:24 <reillyg> reillyg has joined #webmachinelearning
16:05:38 <anssik> q?
16:05:55 <anssik> RRSAgent, draft minutes
16:05:56 <RRSAgent> I have made the request to generate https://www.w3.org/2026/04/02-webmachinelearning-minutes.html anssik
16:07:32 <anssik> s/participants to/participants joining
16:09:28 <anssik> s/or specific/or ask for specific
16:09:49 <anssik> s/capable mode/capable model
16:10:49 <anssik> s/speedy/speed-optimized
16:15:29 <anssik> s/remote-attested -/remote-attested =
16:15:39 <anssik> s/remote -/remote =
16:16:26 <anssik> s/this groups/this group's
16:17:38 <iank_> iank_ has joined #webmachinelearning
16:18:02 <anssik> s/would help/would help enhance the explainer
16:20:01 <anssik> s/may concern/main concern
16:21:54 <anssik> s/also/it can say
16:22:17 <anssik> s/the questions/the question
16:22:31 <anssik> s/should take/should we take
16:23:13 <anssik> s/this is now/WebMCP is now
16:23:51 <anssik> s/capability means/capability it means
16:26:04 <anssik> s/giving parent access/giving the browser access
16:27:41 <anssik> s/we're working/we've worked
16:28:05 <anssik> RRSAgent, draft minutes
16:28:06 <RRSAgent> I have made the request to generate https://www.w3.org/2026/04/02-webmachinelearning-minutes.html anssik
16:28:51 <anssik> Present+ Alex_Nahas
16:29:07 <anssik> Present+ Brandon_Walderman
16:29:22 <anssik> Present+ Ehsan_Toreini
16:29:43 <anssik> Present+ Iris_Johnson
16:30:20 <anssik> Present+ Johann_Hofmann
16:30:39 <anssik> Present+ Julia_Pagnucco
16:30:59 <anssik> Present+ Jean_Vanderdonckt
16:31:20 <anssik> Present+ Saron_Yitbarek
16:31:42 <anssik> Present+ Victor_Huang
16:31:54 <anssik> RRSAgent, draft minutes
16:31:56 <RRSAgent> I have made the request to generate https://www.w3.org/2026/04/02-webmachinelearning-minutes.html anssik
16:32:44 <anssik> Present+ Simone_Onofri
16:32:56 <anssik> Present- simone
16:32:59 <anssik> RRSAgent, draft minutes
16:33:01 <RRSAgent> I have made the request to generate https://www.w3.org/2026/04/02-webmachinelearning-minutes.html anssik
16:37:27 <anssik> s/and AI Agent/and AI Agents
16:37:28 <anssik> RRSAgent, draft minutes
16:37:29 <RRSAgent> I have made the request to generate https://www.w3.org/2026/04/02-webmachinelearning-minutes.html anssik
16:38:15 <anssik> s/Topic: Announcement: Workshop on E-commerce for Humans and AI Agent/Topic: Announcement: Workshop on E-commerce for Humans and AI Agents
16:38:17 <anssik> RRSAgent, draft minutes
16:38:19 <RRSAgent> I have made the request to generate https://www.w3.org/2026/04/02-webmachinelearning-minutes.html anssik
19:00:47 <Zakim> Zakim has left #webmachinelearning