Working group Track C
AI and ML research outputs
The AI and ML research outputs working group defines the documentation patterns for machine-learning artefacts — model cards, system cards, datasheets for datasets, benchmark cards and evaluation-suite descriptors. It coordinates with the Partnership on AI, Hugging Face, MLCommons and NIST so a model released alongside a paper carries the structured metadata that auditors, downstream users and reviewers now demand. The group sits at the intersection of research outputs, reproducibility and AI governance.
Charter and scope
What this working group covers
The AI and ML research outputs working group defines the documentation patterns for machine-learning artefacts — model cards, system cards, datasheets for datasets, benchmark cards and evaluation-suite descriptors. It coordinates with the Partnership on AI, Hugging Face, MLCommons and NIST so a model released alongside a paper carries the structured metadata that auditors, downstream users and reviewers now demand. The group sits at the intersection of research outputs, reproducibility and AI governance.
- Model cards (Mitchell et al., 2019) and their evolution
- System cards and deployment-context documentation
- Datasheets for datasets (Gebru et al.) and dataset nutrition labels
- Benchmark cards and evaluation-suite descriptors
- NIST AI RMF and EU AI Act documentation requirements
- Hugging Face Hub model-card schema alignment
- Provenance, training-data lineage and licence statements
Composition
Current composition
Co-chair
Seat open
Optional second chair, prioritised for regional or sector balance.
Apply for co-chair →Community seats
12 of 12 seats open
Seats are allocated against a published rubric covering domain expertise, institutional diversity, regional balance, and an explicit slot for early-career researchers.
Apply for a community seat →Deliverables
Planned and delivered deliverables
- Model card profile (CASRAI-aligned, Hugging Face compatible)
- Datasheet template aligned with Gebru et al.
- Benchmark card profile
- Cross-walk: NIST AI RMF, EU AI Act, model-card fields
- Repository deposit guidance for ML artefacts
Recent activity
Working-group cadence and milestones
Illustrative working-group cadence for the 2026 forming round. Substantive deliverables and meeting minutes will be linked here as the group convenes.
2026-05-15 · Milestone
Working group forming — call for chair candidates
Open call for chair and co-chair of the AI and ML research outputs working group. Apply by 2026-06-30.
2026-04-22 · Release
Domain scope confirmed for v2026.2
Scope of the AI and ML research outputs domain confirmed against the v2026.2 dictionary release plan.
2026-03-10 · Added
Forming round announced
CASRAI announced the 2026 forming round for all 20 working groups, with seat allocations and review rubric published.
Open consultations
Currently in public comment
Currently none open. Subscribe via /get-involved/comment for future calls.
External bodies
Standards and organisations we work with
Going deeper on CASRAI
Related CASRAI guidance
Adjacent working groups
Related working groups
How to join
Apply to a seat on this working group
All seats on this working group are open for the 2026 forming round. The application is short — name, institution, ORCID iD (optional), the seat you are applying for, and a paragraph on why this domain. Decisions are returned within four weeks of the close of the open-call window.
FAQ
Frequently asked
- What does this working group do?
The AI and ML research outputs working group is the community body that drafts, reviews and ratifies dictionary entries for the AI and ML research outputs domain. It maintains scope, cross-walks to adjacent standards and the cadence of public review. Its remit is summarised on this page; the canonical terms it stewards live at /dictionary/domain/ai-ml-research-outputs.
- Who can apply to join?
Working-group membership is open to qualified practitioners — researchers, research-office staff, librarians, publishers, repository managers, integrity officers, CRIS administrators, regulators and funders. Institutional membership is not required. The qualification test is competence and time, not affiliation. Apply via the working-group application form.
- What is the time commitment?
Cadence: Quarterly virtual sessions; two release windows per year (March and September).. Time commitment averages four hours per quarter for an active member, more for a chair. Asynchronous review happens between meetings via email. Chairs are recognised on the editorial masthead and serve a two-year term, renewable once.
- How are seats allocated?
Each working group has 12 community seats plus a chair and (optionally) a co-chair. Seats are filled by an open call reviewed against a published rubric: domain expertise, institutional diversity, regional balance, and an explicit slot for early-career researchers. All seats are currently open for the 2026 forming round — see how to apply.
- Can my institution sponsor a seat?
Institutional sponsorship is welcome but is not a route to a guaranteed seat. Sponsors support meeting infrastructure, public-comment publication and contributor honoraria; they receive named acknowledgement on the working-group page and in release notes. Sponsorship enquiries go to [email protected]. Seat selection remains on merit.
All working groups
