Time

Agenda Item

Lead

Notes

5 mins

• Welcome & antitrust notice

• Introduction of new members

• Agenda review

• Recording with transcription is now automatic

Chairs

• Antitrust Policy Notice: Attendees are reminded to adhere to the meeting agenda and not participate in activities prohibited under antitrust and competition laws.

• ToIP Policy: Only members of ToIP who have signed the necessary agreements are permitted to participate in this activity beyond an observer role.

• ToIP AIMWG IPR Policy: see WG wiki

5 mins

• Introduction of new members

• Any general announcements news, that could be of interest to the TF

• To review of action items that are not in the agenda below

All

• Any new members to introduce

  • @Jim Mason

  •  

• Any news to share

  • @Steven Milstein shared this news on “harms”3 Key Takeaways From the First Senate Hearing on AI Chatbot HarmsPreview

  • @sankarshan shared this paper from anthropic with very good analysis on confidential computing use with ai agents (Tees etc.) https://cdn.discordapp.com/attachments/1379651804790264018/1420991010648887397/Confidential_Inference_Systems.pdf?ex=68dea8c1&is=68dd5741&hm=e1d8c2b140a5e6ac588801fa51cfc4c301c3e3fb7d28544be655591055f31998&

  • @Nicky Hickman shared her mapping of agentic ai space https://docs.google.com/spreadsheets/d/1S96-vfnrpUiL4OItT4gpYt2nmQqpMUoC/edit?usp=sharing&ouid=102748924597224658467&rtpof=true&sd=true

  • @Eric Drury Introduction | Universal Tool Calling Protocol (UTCP)

    Another framework presented recently to integrate API discovery/calling (multiple protocols and paradigms supported) into AI agents systems.

  • @mitchell Travers this is recent Intel and AMD trusted enclaves, a foundation for network security, fall to physical attacks

  • Open

    

     

  • Who is attending IIW week (Oct 20-24) in person? Presentations?

50 mins

Organize WG’s work in defined threads/initiatives.

• MAY become a TF, but not necessary (if not a TF, we will have standing agenda items for each in the WG call).

• Better defined goals and deliverables

• Fit the AIMWG charter

• Complimentary relationships

ALL

@Wenjing Chu

The names are placeholders. Please add names who would like to help lead the initiatives.

• Anything else to include in the top level to better define & organize these initiatives within the WG?

• Currently active/proposed initiatives are as follows.

1. AI Agent Protocols over TSP @Wenjing Chu

   1. Goal: apply TSP and ToIP stack (technical and gov) to ai agent protocols to help strengthen trust relationship between ai agents and people.

   2. Deliverables: (1) a protocol specification, (2) an experimental implementation (3) protocol integration: mcp, a2a, … (4) demo apps (5) analysis of security and trust assurances

   3. Plan: 2H 2025 for initial draft/prototype

   4. Fit the AIMWG charter: apply ToIP to AI agents

   5. Its design will leverage TSP, DID, VC, and and human trust principles that can come out of 2, 3, 4 work threads. It can be used in other initiatives as technical means, including other WGs in or outside of ToIP. It’s of course closely related to AI agent protocols e.g. MCP, A2A, ANP, etc.

@Wenjing Chu commented that part of this work was creating a meta-pattern that showed how to apply TSP to any protocol.

2. Use Cases to illustrate trust relationships between AI agents and Humans @Chi Hwa Tang @Neil Thomson

   1. Goals

   2. Deliverables

   3. Plan: may be also be part of the trust framework TF

   4. Fit the AIMWG charter

   5. Complimentary Relationships

3. Trustworthy AI Agent framework TF @Eric Drury

   1. Goals: Introduce a broad and easy-to-navigate trust framework in order that builders, deployers, and users of AI Agent products and services can assess and signal the trustworthiness of those products and services.

   2. Deliverables: Framework + supporting documentation (Framework navigation guide / White Paper).

   3. Plan: Q4-25: Framework alignment, Information collating and organising; Q1-26: Introduction with White Paper.

   4. Fit the AIMWG charter: Addresses both technical AND human elements

   5. Complimentary Relationships: DIF Agent WG…

   6. TF’s meeting time/chairs/scope: Tuesdays 4pm CEST

@Neil Thomson note Moves to circumvent personal agents for online purchases?

Example: Google has integrated Gemini in Chrome - which will support handling your purchases for any web page in the browser, have you not given up control over your (bank & credit card) accounts?

@mitchell is telling stories!

4. Human Experience aspects of AI and Human Trust @Nicky Hickman

   1. Goals

      1. TBD at Symposium session

   2. Deliverables

   3. Plan:

      1. Consider to be a part of trust framework task force

   4. Fit the AIMWG charter

   5. Complimentary Relationships

0 mins

Action Items

ALL

1. List action items to follow up

   1. AP2

   2. Update the WG wiki page with the introduction of the active work threads

   3.