Model Card
What is Model Card?
A documentation framework for machine learning models that provides details about a model's intended use, performance metrics, training data, limitations, and ethical considerations. Model cards promote transparency and responsible AI use.
Terms that appear alongside model card
Each of these is named in at least one of the same controls as model card. The number is how many controls name both.
- nist 4 shared controls
- red team 3 shared controls
- eu ai act 3 shared controls
- stress testing 2 shared controls
- incident reporting 2 shared controls
- sandboxing 2 shared controls
- content filtering 2 shared controls
- policy 2 shared controls
Frameworks that govern model card
What the standards actually require on model card
Requirements naming model card across 6 standards, quoted from the control text.
Transparency (Toumeisei 透明性) is the sixth of 10 Principles per Japan AI Guidelines for Business + reinforced by Hiroshima AI Process Code of Conduct (October 2023) which establishes voluntary transparency commitments for frontier AI developers.
JP-AIG-Transparency-Documentation-Model-Card-System-Card-Tier-Based-Disclosure-Hiroshima-Code-of-Conduct · Japan AI Guidelines Transparency + Documentation + Model Card + System Card + Datasheet + Tier-Based Disclosure + Hiroshima Code of Conduct + AI Generated Content + Watermarking + C2PA + User Notification →Operations + Lifecycle cover IEEE 7000 in deployed system operation. Per public IEEE 7000-2021 abstract + Wikipedia + academic literature (full IEEE text NOT reproduced): monitoring in operation including: ethical KPIs continuous tracking + drift monitoring (c...
IEEE7000-Operations-Lifecycle-OngoingMonitoring-Incident-Decommissioning · IEEE 7000 - Operations + Lifecycle + Ongoing AI Risk Monitoring + Data Provenance + Retention + Privacy + Safe Deployment + Decommissioning + Disposal →The developer must make available, to the extent feasible, documentation through artifacts such as model cards, dataset cards or impact assessments necessary for the deployer (or its third party) to complete an impact assessment under 6-1-1703(3).
CO-AIA-1702-3 · Impact-Assessment Support Artifacts →Security Dimension 6 Data Integrity per X.805 Clause 6.6: Data Integrity ensures the correctness or accuracy of data. The data is protected against unauthorized modification + deletion + creation + replication and provides an indication of these unauthorized a...
X805-Dim6-Data-Integrity-Correctness-Accuracy-Unauthorized-Modification-Deletion-Detection · ITU-T X.805 Security Dimension 6 - Data Integrity + Correctness + Accuracy + Unauthorized Modification Prevention + Deletion Detection + Hashing + HMAC + Digital Signatures + Merkle Trees + Blockchain Integrity + File Integrity Monitoring (FIM) →Adhere to OECD AI Principles Section 1.3 (Transparency and explainability). AI actors should commit to (a) transparency and responsible disclosure regarding AI systems to foster understanding of AI systems + make stakeholders aware of their interactions with A...
OECDAI-2 · Transparency, Explainability, and Public-Facing Disclosure →Operate LLM governance + inventory + risk + change management. Requirements include (a) maintain inventory of LLM systems + models + datasets + tools + agents + integrations + (b) maintain risk assessment + classification per LLM system aligned to applicable r...
OWASPLLM-7 · LLM Governance, Inventory, Risk and Change Management →Questions people ask about model card
What is Model Card?
Why is Model Card important for compliance?
Which compliance frameworks address Model Card?
Where can I learn more about Model Card?
See how Model Card applies across compliance frameworks
Our platform maps 686 frameworks with 311K cross-framework control mappings. Explore how this concept is addressed across standards.