Trust KR Actions on DIF Delegated Authority Evaluations

A model performance plan for the suggested actions arising from the mapping of the DIF Delegated Authority Evaluations draft to the AI KR CG trust work.
Other Information
Source: "Note on the DIF Delegated Authority Evaluations" (DIFdelegateauth.md), a working note for the Trust KR task force in the w3c-cg/aikr GitHub repository, open for comments, prepared 20 November 2026. The note maps the DIF Delegated Authority Evaluations draft of 17 November 2026 (v1.0-rc1, Decentralized Identity Foundation, Trusted AI Agents Working Group) to the AI KR CG trust work, led by the knowledge representation of trust note (kroftrust.md), and to the AI KR CG landscape crosswalks, including the Agent Trust Protocols and Specifications Explorer and the cross-protocol use cases. This plan renders the note's "Suggested actions for the task force" as Goals and Objectives, with stakeholder roles and performance indicators; the note itself does not group them into goals.
Cross-references (all drafts): Draft report on the knowledge representation of trust: https://w3c-cg.github.io/aikr/trust/kroftrust.html | Agent Trust Protocols and Specifications Explorer (mapping ARIA, did:atp and other agent trust and identity specifications by coverage and maturity): https://w3c-cg.github.io/aikr/trust/trustinitiativeexplorer.html | W3C AI KR Trust Record visualiser (Mappings tab): https://w3c-cg.github.io/aikr/trust/AI_Agent_Trust_Record.html#mappings | W3C AI KR Trust Record checker: link to be added (per the note) | Case library of cross-protocol use cases: https://github.com/w3c-cg/aikr/blob/main/trust/trustusecases.md | Trust KR task force: https://github.com/w3c-cg/aikr/blob/main/trust/trust-kr-taskforce.md
DIF report: Delegated Authority Evaluations, Decentralized Identity Foundation (DIF). Editors: Dmitri Zagidulin (Interop Alliance) and Juan Caballero (learningProof UG). Status: Working Group Consensus Draft, v1.0-rc1, 17 November 2026, pending DIF Steering Committee approval; verdicts may change before the report is approved. A qualitative evaluation of authorization and access control specifications through the lens of agentic permissions and delegation, giving each specification a Yes, Partial or No verdict per criterion with narrative reasoning. The report does not itself specify a protocol and describes itself as synoptic, not exhaustive. Companion documents: Problem Space Report, Threat Model, Governance of Delegated Authority report, and an agentic-delegation user-story walk-through (video).
Specifications evaluated: OAuth 2.0 (with RFC 9700) and the OAuth 2.1 draft; AAuth (draft-hardt-oauth-aauth-protocol); the certificate capability family (zCaps, UCANs, KYA-OS); W3C Verifiable Credentials as a way to model chained delegation; Cedar and Janssen Cedarling; AuthZEN; with dSD-JWT and AuthZed/SpiceDB in an appendix.
Criteria: seven required (1 Attenuated delegation; 2 Accountability, distinguishing the agent from its principal or operator; 3 Confused-deputy resistance; 4 Representability of authorization policies: functional, re-delegation, advisory and disclosure-governance rules; 5 Chainability; 6 Cross-organizational, locally verifiable execution; 7 Self-revocability), plus 8 Proof of possession as a nice-to-have.
Merged scorecard, as published (order: OAuth, AAuth, zCaps, UCANs, KYA-OS, Cedar, AuthZEN, dSD-JWT; Y = Yes, P = Partial, N = No): 1 Attenuated: P, P, Y, Y, Y, P, N, N | 2 Accountable: Y, Y, P, P, Y, Y, P, P | 3 Confused deputy: P, Y*, Y, Y, Y, N, N, P | 4 Policies: P, P, P, P, P, Y, P, N | 5 Chainable: P, Y, Y, Y, Y, N, N, Y | 6 Cross-org / local: Y, Y, Y, Y, P, P, N, Y | 7 Self-revocable: P, P, Y, Y, P, P, P, N | 8 Proof of possession: Optional, Y, Y, Y, Y, N, N, Y. *AAuth is Yes in three of its four modes; the identity-based mode is ambient authority. W3C Verifiable Credentials receive no per-criterion verdicts: a VC has no parent, chain or attenuation rule, so chaining must be modelled in the payload; the report places VCs at the root of a chain, binding the operator to the first delegation.
Main findings of the DIF report: The specifications fall into three families: certificate capabilities (zCaps, UCANs, KYA-OS), server-mediated authorization (OAuth, AAuth) and policy decision systems (Cedar, AuthZEN). The capability family is the only one considered viable as the backbone of the authority flow, since it alone meets the chain criteria offline and holder-side. No specification is complete. Shared gaps: no delegation-control policy vocabulary (e.g. "do not delegate further"); revocation online-only in every surveyed mechanism, with no defined semantics under network partition; delegation-chain privacy left as future work. Recommended direction: compose a four-layer delegation stack (identity as entry point, certificate chain for the authority flow, policy and caveats for restrictions, signed presentation for possession); near-term work is composition, with no need for a competing specification; AuthZEN alone should not be the basis for access decisions in a delegation stack, since it is ambient authority.
Overlap: The DIF report evaluates delegation and authorization specifications; the AI KR CG explorer covers agent trust and identity protocols such as ARIA and did:atp. AuthZEN is the only specification in both. Shared ground (per the note): attenuated delegation and the AI KR authority-continuity model; agent-versus-operator accountability and the Trust Record parties (author, operator, authoriser, owner, funder); rejecting caveats the verifier cannot evaluate and the onUnknown relying-party policy; fail-closed revocation with a freshness verdict and the three-valued evidence results (pass, fail, could-not-determine) and stalenessBound; and the four-layer delegation stack and four of the Trust Record field groups. The note calls the overlap enough for an exchange in both directions.
Proposed mapping of DIF criteria to W3C AI KR Trust Record field groups (for discussion, not from the report): 1 Mandate and authority; 2 Parties; 3 Mandate and authority, Policy; 4 Policy; 5 Mandate and authority; 6 Interface and keys, Evidence; 7 Lifecycle; 8 Interface and keys. Use-case links: UC-01 (action exceeds authority) tests criterion 1; UC-02 (revocation with cached evidence) tests criterion 7; UC-05 (pseudonymous or pairwise identifier) matches the report's recommendation of ephemeral or pairwise identifiers per delegation; UC-07 (copied or replayed credential) tests criterion 8; UC-09 (hiring agent) touches criterion 2 and the agent-versus-operator gap.
AIKR work with no counterpart in the DIF report: resolution dependency and anchoring path (ledger, DNS, DID), including the DNS/DID note and the root KSK rollover; the evidence scale for specifications (documented, author-published, independently reproduced); notes on Base L2, "Trust Without Trusting" and the Agent Identity Protocol; trust in non-agentic AI (trustinai(non_agentic).md).
Comments: open an issue on w3c-cg/aikr with "DIF note" (or the file name) in the title, or post to public-aikr@w3.org with [trust-dif] in the subject line. The latest commit of the note visible in the source copy is a7f373e by GitHub user Starborn ("Update DIFdelegateauth.md"); the note names no author.
Details not specified in the source: The note names the Trust KR task force as the group for which the actions are suggested, but does not assign individual owners, start or target dates, priorities or sequencing among the actions, how completion will be verified or reported, or what response is sought from DIF. These are left blank or marked TBD in this rendition, for the performers to determine as they see fit. Targets recorded here are only those the note itself implies: one status per action, two kinds of limit (re-delegation and depth), two report editors, and two groups (ARIA and ATP). All ActualResults are TBD. Roles: the schema's RoleType allows only Performer and Beneficiary. Parties outside the task force whose action is needed (the report's editors, Alan Karp, and the ARIA and ATP groups) are shown as Performers whose participation is at their discretion; DIF and its Trusted AI Agents Working Group are shown as Beneficiaries of the input offered under Goal 3. People the note identifies as working in both bodies of work (Nicola Gallo, Victor Lu) are listed without roles, since the note assigns them none.
Submitter's Note: This StratML Part 2 rendition has been compiled from the source by Claude.ai and can be edited in the form at https://stratml.us/forms/Claude/Part2.html to flesh out the stakeholder roles and performance indicator data as it becomes available.
The vision, mission, and values statements have been inferred by Claude at my request.

🏢W3C Artificial Intelligence Knowledge Representation Community Group

AI KR CG

W3C community group whose trust work, led by the knowledge representation of trust note, is carried forward by its Trust KR task force.

Stakeholders:
👥Trust KR task force: Task force of the AI KR CG working on the knowledge representation of trust; the group for which the note's suggested actions are proposed. Its contributors are listed at https://github.com/w3c-cg/aikr/blob/main/trust/trust-kr-taskforce.md
Implementer: Carries out the suggested action. The note names no individual owners within the task force. (Performer)

Source & Submitter

Source:https://github.com/w3c-cg/aikr/blob/main/trust/DIFdelegat...

Period: to

Published:


👤

Given name:Owen

Surname:Ambur

Email:Owen.Ambur@verizon.net

Vision

Agent trust and delegation work across communities is mutually informed and readily compared.

Mission

To map related agent trust and delegation work to the AI KR CG trust work, close the gaps the mapping reveals, and offer AIKR findings in return.

Values

Reciprocity: Contribute as well as borrow. Based on: "AI KR input on the DIF future-work gaps… and the DIF criteria to fill the AI KR gaps."
Openness: Invite comment from all parties through public channels. Based on: the invitations to the task force, the DIF editors and the ARIA and ATP groups, via GitHub issues or public-aikr@w3.org.
Accuracy: State qualifications and uncertainties plainly. Based on: "matches names, not confirmed identities," the warning that DIF's verdicts may change before approval, and the mappings being "proposed for discussion."
Evidence: Distinguish documented claims from independently reproduced ones. Based on: the explorer's evidence scale (documented, author-published, independently reproduced).

Goal 1: Landscape🔗

Reflect the DIF Delegated Authority Evaluations report in the AI KR CG trust landscape resources.
Stakeholders:
👥Trust KR task force: Task force of the AI KR CG working on the knowledge representation of trust; the group for which the note's suggested actions are proposed. Its contributors are listed at https://github.com/w3c-cg/aikr/blob/main/trust/trust-kr-taskforce.md
Implementer: Carries out the suggested action. The note names no individual owners within the task force. (Performer)
Other Information
Landscape ~ The first two suggested actions concern the explorer and the trust note. The note observes that the report covers authorization and delegation specifications while the explorer covers agent trust and identity protocols; overlap is limited to AuthZEN, and ARIA and did:atp are not among the evaluated specifications. The report is a qualitative review by its editors; the explorer adds an evidence scale (documented, author-published, independently reproduced).

Objective 1.1: Explorer Entry🔗

Add the DIF Delegated Authority Evaluations report to the explorer's corollary initiatives, with its status and date.
Stakeholders:
👥Trust KR task force: Task force of the AI KR CG working on the knowledge representation of trust; the group for which the note's suggested actions are proposed. Its contributors are listed at https://github.com/w3c-cg/aikr/blob/main/trust/trust-kr-taskforce.md
Implementer: Carries out the suggested action. The note names no individual owners within the task force. (Performer)
Other Information
Explorer: https://w3c-cg.github.io/aikr/trust/trustinitiativeexplorer.html. Status and date per the note: Working Group Consensus Draft, v1.0-rc1, 17 November 2026, pending DIF Steering Committee approval; verdicts may change before approval.

Performance Indicators

1.1.1: Explorer Entry
TypeStart DateEnd DateStatusDescription
TargetAddedThe report is listed among the explorer's corollary initiatives, with its status and date.
ActualTBDTo be reported by the performers.

Objective 1.2: Trust Note Entry🔗

Update the DIF entry in the trust note's landscape section.
Stakeholders:
👥Trust KR task force: Task force of the AI KR CG working on the knowledge representation of trust; the group for which the note's suggested actions are proposed. Its contributors are listed at https://github.com/w3c-cg/aikr/blob/main/trust/trust-kr-taskforce.md
Implementer: Carries out the suggested action. The note names no individual owners within the task force. (Performer)
Other Information
Trust note: draft report on the knowledge representation of trust (kroftrust.md), https://w3c-cg.github.io/aikr/trust/kroftrust.html

Performance Indicators

1.2.1: Trust Note Entry
TypeStart DateEnd DateStatusDescription
TargetUpdatedThe DIF entry in the trust note's landscape section reflects the report.
ActualTBDTo be reported by the performers.

Objective 1.3: Criteria View🔗

Determine whether to add the DIF criteria as a view in the explorer, using the note's mappings once they are agreed.
Stakeholders:
👥Trust KR task force: Task force of the AI KR CG working on the knowledge representation of trust; the group for which the note's suggested actions are proposed. Its contributors are listed at https://github.com/w3c-cg/aikr/blob/main/trust/trust-kr-taskforce.md
Implementer: Carries out the suggested action. The note names no individual owners within the task force. (Performer)
Other Information
Criteria View ~ The note says "Consider adding" the view; the mappings are proposed for discussion and not yet agreed. Proposed mapping of DIF criteria to Trust Record field groups: 1 Attenuated delegation - Mandate and authority; 2 Accountability (agent vs operator) - Parties; 3 Confused-deputy resistance - Mandate and authority, Policy; 4 Representability of policies - Policy; 5 Chainability - Mandate and authority; 6 Cross-org, locally verifiable - Interface and keys, Evidence; 7 Self-revocability - Lifecycle; 8 Proof of possession - Interface and keys.

Performance Indicators

1.3.1: Decision
TypeStart DateEnd DateStatusDescription
TargetDecidedA decision is made on whether to add the DIF criteria as a view in the explorer.
ActualTBDTo be reported by the performers.
Other Information
The note says "Consider adding" the view, using the mappings once agreed; the decision depends on agreement on the proposed mappings, for which the note sets no process.

Goal 2: Gaps🔗

Close the gaps on the AIKR side that the DIF report raises and the AIKR trust work does not yet treat.
Stakeholders:
👥Trust KR task force: Task force of the AI KR CG working on the knowledge representation of trust; the group for which the note's suggested actions are proposed. Its contributors are listed at https://github.com/w3c-cg/aikr/blob/main/trust/trust-kr-taskforce.md
Implementer: Carries out the suggested action. The note names no individual owners within the task force. (Performer)
Other Information
Gaps ~ The note's mapping uses three statuses: aligned (both cover it), AIKR adds (AIKR covers ground the DIF report does not score), and gap (the DIF report raises it, the AIKR trust work does not yet treat it). DIF work that goes further: confused-deputy resistance, proof of possession, and limits on re-delegation and chain depth, none yet treated in the AI KR trust work. Delegation-chain privacy is open on both sides (DIF lists it as future work; UC-05 covers pairwise identifiers, not chain privacy) and is not among the suggested actions.

Objective 2.1: Confused-Deputy Resistance🔗

Add confused-deputy resistance to the trust decision or the W3C AI KR Trust Record.
Stakeholders:
👥Trust KR task force: Task force of the AI KR CG working on the knowledge representation of trust; the group for which the note's suggested actions are proposed. Its contributors are listed at https://github.com/w3c-cg/aikr/blob/main/trust/trust-kr-taskforce.md
Implementer: Carries out the suggested action. The note names no individual owners within the task force. (Performer)
Other Information
DIF criterion 3. Current AIKR status per the note: Gap; no dedicated treatment. Closest: the trust decision's authority and enforcement-location dimensions. Proposed Trust Record field groups: Mandate and authority, Policy.

Performance Indicators

2.1.1: Treatment
TypeStart DateEnd DateStatusDescription
TargetAddedConfused-deputy resistance is treated in the trust decision or the W3C AI KR Trust Record.
ActualTBDTo be reported by the performers.

Objective 2.2: Proof of Possession🔗

Add proof of possession to the trust decision or the W3C AI KR Trust Record.
Stakeholders:
👥Trust KR task force: Task force of the AI KR CG working on the knowledge representation of trust; the group for which the note's suggested actions are proposed. Its contributors are listed at https://github.com/w3c-cg/aikr/blob/main/trust/trust-kr-taskforce.md
Implementer: Carries out the suggested action. The note names no individual owners within the task force. (Performer)
Other Information
DIF criterion 8 (nice-to-have). Current AIKR status per the note: Gap for the mechanism; the use case (UC-07, copied or replayed credential) is in place. Not defined in the trust note; related: the key publisher field (who publishes the key set a verifier trusts) and key-to-domain binding checks. The DIF report recommends a separate per-request proof whose key matches the leaf delegatee. Proposed Trust Record field group: Interface and keys.

Performance Indicators

2.2.1: Treatment
TypeStart DateEnd DateStatusDescription
TargetAddedProof of possession is treated in the trust decision or the W3C AI KR Trust Record.
ActualTBDTo be reported by the performers.

Objective 2.3: Delegation Limits🔗

Add re-delegation and depth limits to the authority-continuity cases.
Stakeholders:
👥Trust KR task force: Task force of the AI KR CG working on the knowledge representation of trust; the group for which the note's suggested actions are proposed. Its contributors are listed at https://github.com/w3c-cg/aikr/blob/main/trust/trust-kr-taskforce.md
Implementer: Carries out the suggested action. The note names no individual owners within the task force. (Performer)
Other Information
Relates to DIF future-work gap 1 (a standard policy and caveat vocabulary for delegation control: time windows, re-delegation and depth limits). Current AIKR status per the note: Gap; the authority-continuity model (scope may narrow but not widen without explicit authorisation, with four proposed cases) covers narrowing but has no vocabulary for re-delegation or depth limits. The note's mapping table also suggests the DIF per-hop attenuation invariant could be added to the authority-continuity cases.

Performance Indicators

2.3.1: Limits
TypeStart DateEnd DateNumberStatusDescription
Target2AddedTwo kinds of limit, re-delegation limits and depth limits, are added to the authority-continuity cases.
ActualTBDTo be reported by the performers.
Other Information
Together with the indicators for Objectives 2.1 and 2.2, this covers the three gaps the note asks the task force to close. Part 2 carries performance indicators at the Objective level only, so no Goal-level count is recorded.

Goal 3: DIF Input🔗

Contribute AIKR terms and semantics as input on the DIF future-work gaps on revocation and policy exchange.
Stakeholders:
👥Trust KR task force: Task force of the AI KR CG working on the knowledge representation of trust; the group for which the note's suggested actions are proposed. Its contributors are listed at https://github.com/w3c-cg/aikr/blob/main/trust/trust-kr-taskforce.md
Implementer: Carries out the suggested action. The note names no individual owners within the task force. (Performer)
🏢Decentralized Identity Foundation: Publisher of the Delegated Authority Evaluations draft, pending approval by its Steering Committee.
Recipient: Prospective recipient of the AIKR input on its future-work gaps; whether and how DIF uses it is DIF's decision. (Beneficiary)
👥DIF Trusted AI Agents Working Group: The group listed as producing the delegated authority reports.
Recipient: The group listed as producing the delegated authority reports, whose future-work gaps the AIKR input addresses. (Beneficiary)
👤Nicola Gallo: Member of DIF Trusted AI Agents Working Group; co-chair. Also an AI KR CG participant.
Other Information
DIF Input ~ The suggested action is to offer the AIKR resolution-dependency terms (stalenessBound, onUnknown, coupling) and comparison semantics to DIF as input on its future-work gaps on revocation under partition and cross-organizational policy exchange. The note pairs resolution dependency with the revocation gap and comparison semantics with the policy-exchange gap; this plan splits them accordingly. DIF future-work gaps, in the report's order: (1) standard policy and caveat vocabulary for delegation control; (2) an advisory tier for requests recorded but not enforced; (3) agent-versus-operator semantics and identity binding that survive cross-organizational, multi-hop delegation; (4) offline and decentralized revocation under network partition; (5) cross-organizational policy exchange with agreed semantics. People in both bodies of work are listed in the note; it compares named authors and chairs only, since the DIF working group's membership is not public, and matches names, not confirmed identities.

Objective 3.1: Revocation Semantics🔗

Offer the AIKR resolution-dependency terms (stalenessBound, onUnknown, coupling) to DIF as input on its future-work gap on revocation under network partition.
Stakeholders:
👥Trust KR task force: Task force of the AI KR CG working on the knowledge representation of trust; the group for which the note's suggested actions are proposed. Its contributors are listed at https://github.com/w3c-cg/aikr/blob/main/trust/trust-kr-taskforce.md
Implementer: Carries out the suggested action. The note names no individual owners within the task force. (Performer)
Other Information
DIF future-work gap 4. DIF finds revocation online-only in every surveyed mechanism, with no defined semantics under network partition. AIKR counterpart per the note: resolution dependency as a PACELC position (coupled vs decoupled verification paths), stalenessBound (longest window a revoked or superseded artefact can still verify), onUnknown, coupling. The note says this adds a vocabulary for stating how a verifier behaves when it cannot reach the revocation source, and is worth offering to DIF. Related alignment: DIF's fail-closed revocation checks with a freshness verdict correspond to the three-valued evidence results (pass, fail, could-not-determine).

Performance Indicators

3.1.1: Offer
TypeStart DateEnd DateStatusDescription
TargetOfferedThe resolution-dependency terms (stalenessBound, onUnknown, coupling) are offered to DIF as input on revocation under network partition.
ActualTBDTo be reported by the performers.
3.1.2: DIF Response
TypeStart DateEnd DateStatusDescription
TargetTBDThe note sets no target for this result; any target is left to the performers to determine.
ActualTBDTo be reported by the performers.

Objective 3.2: Policy Exchange🔗

Offer AIKR comparison semantics to DIF as input on its future-work gap on cross-organizational policy exchange with agreed semantics.
Stakeholders:
👥Trust KR task force: Task force of the AI KR CG working on the knowledge representation of trust; the group for which the note's suggested actions are proposed. Its contributors are listed at https://github.com/w3c-cg/aikr/blob/main/trust/trust-kr-taskforce.md
Implementer: Carries out the suggested action. The note names no individual owners within the task force. (Performer)
Other Information
DIF future-work gap 5. AIKR counterpart per the note: comparison semantics (term-level crosswalks fail behaviourally; canonical-form version and declared semantics; vocabulary crosswalk). The note says this adds evidence that agreed terms are not enough without declared comparison semantics.

Performance Indicators

3.2.1: Offer
TypeStart DateEnd DateStatusDescription
TargetOfferedAIKR comparison semantics are offered to DIF as input on cross-organizational policy exchange.
ActualTBDTo be reported by the performers.
3.2.2: DIF Response
TypeStart DateEnd DateStatusDescription
TargetTBDThe note sets no target for this result; any target is left to the performers to determine.
ActualTBDTo be reported by the performers.

Goal 4: Review🔗

Validate the mapping between the DIF report and the AIKR trust work.
Stakeholders:
👥Trust KR task force: Task force of the AI KR CG working on the knowledge representation of trust; the group for which the note's suggested actions are proposed. Its contributors are listed at https://github.com/w3c-cg/aikr/blob/main/trust/trust-kr-taskforce.md
Implementer: Carries out the suggested action. The note names no individual owners within the task force. (Performer)
Other Information
Review ~ The note asks: Is the mapping right? Is anything missing or misread? Comments are invited from the Trust KR task force, the DIF report's editors and contributors, and the ARIA and ATP groups, via an issue on w3c-cg/aikr with "DIF note" in the title or a post to public-aikr@w3.org with [trust-dif] in the subject line. None of the editors or named contributors of the evaluation draft appears in the AI KR CG participant list.

Objective 4.1: Editor Comments🔗

Obtain comments from the DIF report's editors on the mapping and on trustusecases.md.
Stakeholders:
👥Delegated Authority Evaluations editors: Editors of the DIF Delegated Authority Evaluations report.
Reviewer: Invited to comment on the mapping and on trustusecases.md; participation is at their discretion. (Performer)
👤Dmitri Zagidulin: Member of Delegated Authority Evaluations editors (Interop Alliance).
Reviewer: Invited, as an editor of the report, to comment; participation is at his discretion. (Performer)
👤Juan Caballero: Member of Delegated Authority Evaluations editors (learningProof UG).
Reviewer: Invited, as an editor of the report, to comment; participation is at his discretion. (Performer)
👤Alan Karp: SitePassword. Contributed to the evaluation draft and the companion threat model, and edits the Governance of Delegated Authority report. DIF describes his access management use cases as a key conceptual foundation for the Trusted AI Agents WG approach. The note says his view on the case library would be especially welcome. He does not appear in the AI KR, ATP or ARIA CG participant lists.
Reviewer: Invited to comment; the note says his view on the case library would be especially welcome. Participation is at his discretion. (Performer)
Other Information
Case library: https://github.com/w3c-cg/aikr/blob/main/trust/trustusecases.md. The note's general invitation also extends to the report's contributors, with Alan Karp's view on the case library singled out.

Performance Indicators

4.1.1: Invitation
TypeStart DateEnd DateStatusDescription
TargetInvitedThe report's editors are invited to comment on the mapping and on trustusecases.md.
ActualTBDTo be reported by the performers.
4.1.2: Editors
TypeStart DateEnd DateNumberStatusDescription
Target2CommentedBoth editors named in the report, Dmitri Zagidulin and Juan Caballero, have commented.
ActualTBDTo be reported by the performers.
Other Information
The note also invites comments from the report's contributors, especially Alan Karp on the case library, but sets no count for them.

Objective 4.2: Self-Assessments🔗

Enable ARIA and ATP entries to be read next to the published DIF scorecard, through self-assessment by those groups against the seven DIF criteria.
Stakeholders:
👥ARIA CG: Group responsible for ARIA, an agent trust and identity specification covered by the AI KR CG explorer but not evaluated in the DIF report.
Self-Assessor: Invited to self-assess against the seven DIF criteria; participation is at the group's discretion. (Performer)
👥ATP CG: Group associated with ATP; did:atp is covered by the AI KR CG explorer but not evaluated in the DIF report.
Self-Assessor: Invited to self-assess against the seven DIF criteria; participation is at the group's discretion. (Performer)
👤Victor Lu: Member of ATP CG (participant) and AI KR CG participant; an author of the companion Governance of Delegated Authority report.
Other Information
Self-Assessments ~ The suggested action is to invite the ARIA and ATP groups to self-assess against the seven DIF criteria, so their entries can be read next to the published scorecard.

Performance Indicators

4.2.1: Invitation
TypeStart DateEnd DateStatusDescription
TargetInvitedThe ARIA and ATP groups are invited to self-assess against the seven DIF criteria.
ActualTBDTo be reported by the performers.
4.2.2: Groups
TypeStart DateEnd DateNumberStatusDescription
Target2Self-assessedBoth groups have self-assessed, so their entries can be read next to the published scorecard.
ActualTBDTo be reported by the performers.