DataCardGenerator

3 inputs 5 sections worked out 4 method steps Checked 2026-10-08

Read the method

Dataset documentation, from the texts themselves

Data card generator: outline the card a buyer or a regulator will read

We read 6 data card standards and legal templates field by field, 11 texts in all, read up to 2026-10-08. Pick where the dataset is used and the format you publish in, and the outline lists what each text asks the card to contain.

Source: What each data card standard asks a dataset to document: every row, with its source

Your figures

Your document

The text the card answers to

California AB 2013 and the EU AI Act Article 53(1)(d) training-content summary

  1. Section: sources

    AB 2013 item (1) asks for the sources or owners of the datasets

  2. Section: personal information

    Open: find out before publishing, since AB 2013 and the Data Cards Playbook both ask

  3. Section: EU training-content summary

    Follow the AI Office template, which the Commission calls a common minimal baseline for the public summary

  4. Format of the card

    A README.md in the dataset repo, rendered as the card, with a YAML metadata section

Get this checked by a specialist

A buyer will ask for the card before the licence. Tell us about the dataset and we reply with what the texts on the board ask of it.

Has a buyer already made the data-holding company an offer for its data?

Does the data-holding company already license or sell this data?

Next: where to reach you. Free, no obligation.

The assumptions

  1. 01

    California AB 2013's items are quoted from the bill text: item (1) the sources or owners of the datasets, item (7) whether they include personal information.

    Source: California AB 2013 bill text, read for the data card generator

  2. 02

    The EU line follows Article 53(1)(d) as the Commission's AI Act Service Desk gives it: a sufficiently detailed summary of the content used for training, made publicly available according to a template provided by the AI Office. What the template is for is quoted from the Commission's explanatory notice.

    Source: European Commission AI Act Service Desk: Article 53, read for the data card generator

  3. 03

    Format lines follow each format's own documentation: the Hugging Face Hub docs, the Croissant specification, the Datasheets paper and the Data Cards paper.

    Source: Hugging Face Hub docs: dataset cards, read for the data card generator

  4. 04

    The outline lists sections; it does not say whether a finished card meets any law.

This site summarises published standards and laws on dataset documentation; it is not legal advice, and no standards body endorses it. If a company licenses a dataset through an introduction we make, the company pays our fee of 12.5% of the deal with a minimum as its exclusive agent, or about 10% where the dataset joins a bundle or the pool, agreed in writing first; any fee a buyer pays us is credited to the company.

Data card standards: law or voluntary, and what each asks about the data

Last updated

One row per standard or legal template, each answer quoted or summarised from its own text.

Data card standard or templateLaw or voluntaryAsks where the data came fromAsks about personal dataSource
California AB 2013a legal obligationasks for the data's sources or provenanceasks whether personal data is includedCalifornia AB 2013 (its own page, listed on this site's dataset page)
Croissant (MLCommons)a voluntary standard or formatasks for the data's sources or provenancenot statedCroissant (MLCommons) (its own page, listed on this site's dataset page)
Data Cards (Google Research)not statedasks for the data's sources or provenanceasks whether personal data is includedData Cards (Google Research) (its own page, listed on this site's dataset page)
Datasheets for Datasetsa voluntary standard or formatasks for the data's sources or provenancenot statedDatasheets for Datasets (its own page, listed on this site's dataset page)
EU AI Act training-content summary (Art. 53(1)(d) and template)a legal obligationnot statednot statedEU AI Act training-content summary (Art. 53(1)(d) and template) (its own page, listed on this site's dataset page)
Hugging Face dataset carda voluntary standard or formatnot statednot statedHugging Face dataset card (its own page, listed on this site's dataset page)
Method and citation

Each standard's own official texts, templates and specifications were read through estate-fetch and stored whole (sha256 on every row). For every text, four questions were answered from its own words: binding or voluntary, asks where the data came from, asks about personal data, what it asks the card to contain. Every stated answer carries the sentence it rests on, checked against the stored text; 'not stated' means the text read does not address the question. A standard with several texts reads 'varies' where they disagree. 6 standards, 11 texts.

“Data card standards: law or voluntary, and what each asks about the data”, Data Card Generator, https://datacardgenerator.com/ (updated 2026-10-08).

Ask about documenting a dataset you plan to license

A buyer will ask for the card before the licence. Tell us about the dataset and we reply with what the texts on the board ask of it.

Has a buyer already made the data-holding company an offer for its data?

Does the data-holding company already license or sell this data?

Tick the box above to send your enquiry.

Guides

The longer answers, with sources.

Questions, answered directly

What is a data card?

A short document that travels with a dataset and says where it came from, what is in it and what it is for. Datasheets for Datasets proposes that every dataset be accompanied by one covering motivation, composition, collection process and recommended uses. (What each data card standard asks a dataset to document: every row, with its source)

Is a data card required by law?

2 of the 6 standards and templates on our board are laws: California AB 2013, which asks a developer of a generative AI system for a high-level summary of its datasets, and the EU AI Act, whose Article 53(1)(d) asks providers of general-purpose AI models to make publicly available a sufficiently detailed summary of the content used for training, according to a template provided by the AI Office. Of the others, 3 are voluntary formats, and the paper and the Playbook read for Data Cards do not say whether it is binding. (What each data card standard asks a dataset to document: every row, with its source)

Which standards ask about personal data?

2 of them: AB 2013 asks whether the datasets include personal information as the CCPA defines it, and Google's Data Cards Playbook asks whether sensitive or personal data was collected. The others do not ask in the documents we read. (What each data card standard asks a dataset to document: every row, with its source)

Does this generator write the card for me?

It builds the outline: which sections the texts you are subject to ask for, and the format's own conventions. The answers in each section are yours, and whether a summary satisfies a law is a question for a lawyer. (What each data card standard asks a dataset to document: every row, with its source)

What file is a Hugging Face dataset card?

A README.md in the dataset's repo. The Hub's docs say a dataset repo renders its README.md as the dataset card, and that a YAML section in that file sets the metadata the Hub displays. (What each data card standard asks a dataset to document: every row, with its source)

What is Croissant?

A metadata vocabulary for machine learning datasets, which the MLCommons working group behind it describes as open and community-built. It gives machine-readable descriptions of a dataset's provenance and permissions, and its 1.0 release includes a complete specification of the format. (What each data card standard asks a dataset to document: every row, with its source)

Where does the generator's outline come from?

From the texts on our board: 11 texts, templates and specifications from 6 standards, each read whole and stored, with every answer kept beside the sentence it rests on. (What each data card standard asks a dataset to document: every row, with its source)

Sources

This site's own dataset, every figure quoted from its source and dated:

  1. European Union (official text)
  2. Croissant Format Specification | Croissant site
  3. European Union (official text)
  4. [2204.01075] Data Cards: Purposeful and Transparent Dataset Documentation for Responsible
  5. The Data Cards Playbook - Data Cards Playbook
  6. NIST AI RMF
  7. Croissant - MLCommons
  8. Dataset Cards · Hugging Face

The question this page is here to answer

What does each published dataset-documentation standard or legal template ask a data card to contain?

Pick the market and the format in the outline tool, and it shows the sections each text asks for.

Answered by Data Card Generator, which reads every enquiry this page takes.

Talk to a specialist

Use this on your own site

Embed it anywhere; attribution stays on. Copy this: 

<script src="https://datacardgenerator.com/embed.js" async></script>
Talk to a specialist