AI Model Card.

A standardised document, introduced by Margaret Mitchell and colleagues in 2018, that describes a machine-learning model's intended use, training data, performance, limitations, and ethical considerations. It belongs to the AI & ML Security category of cybersecurity.

A standardised document, introduced by Margaret Mitchell and colleagues in 2018, that describes a machine-learning model's intended use, training data, performance, limitations, and ethical considerations.

Model cards were proposed in the 2018 paper "Model Cards for Model Reporting" by Margaret Mitchell et al. at Google. They standardise transparent reporting for ML models with sections such as intended uses and out-of-scope uses, training and evaluation data, quantitative performance disaggregated by demographic or operational slice, known ethical considerations, and recommendations. Major providers including Google, Hugging Face, OpenAI, Meta, and Microsoft publish model cards for their models, and the format underpins regulatory transparency under the EU AI Act and the US NIST AI Risk Management Framework. From a security perspective, model cards help defenders understand attack surface, data lineage, and known failure modes, and they support due diligence before integrating a third-party model into a product.

Defences for AI Model Card typically combine technical controls and operational practices, as detailed in the full definition above.

Common alternative names include: Model card, ML model card.