{"id":135766,"date":"2026-09-04T17:36:16","date_gmt":"2026-09-04T12:06:16","guid":{"rendered":"https:\/\/www.guvi.in\/blog\/?p=135766"},"modified":"2026-09-04T17:36:17","modified_gmt":"2026-09-04T12:06:17","slug":"model-cards-and-datasheets","status":"publish","type":"post","link":"https:\/\/www.guvi.in\/blog\/model-cards-and-datasheets\/","title":{"rendered":"Model Cards and Datasheets: Documenting AI Responsibly"},"content":{"rendered":"\n<p>Model cards and datasheets are structured documents that describe AI models and datasets in a clear, standardized way. They help teams communicate capabilities, limitations, and ethical considerations to developers, auditors, and end users. As regulations and governance frameworks mature, these documents are becoming essential for responsible AI deployment.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>TL;DR Summary<\/strong><\/h3>\n\n\n\n<ul>\n<li>Model cards document trained models; datasheets document datasets.<\/li>\n\n\n\n<li>Both follow standardized sections to improve transparency and accountability.<\/li>\n\n\n\n<li>They cover intended use, performance metrics, data provenance, and limitations.<\/li>\n\n\n\n<li>Increasingly required by governance frameworks and regulations such as the EU AI Act.<\/li>\n\n\n\n<li>Should be created early, updated regularly, and treated as living documents.<\/li>\n<\/ul>\n\n\n\n<h4 class=\"wp-block-heading\"><strong>Direct Answer<\/strong><\/h4>\n\n\n\n<p><strong>Model cards<\/strong> are short, human-readable documents that summarize a trained model\u2019s intended use, performance, training data, limitations, and ethical considerations. <strong>Datasheets for datasets<\/strong> describe a dataset\u2019s motivation, composition, collection process, preprocessing, distribution, and maintenance. Together, they provide transparency about how an AI system was built, what it can and cannot do, and how it should be used responsibly.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\"><strong>What Are Model Cards?<\/strong><\/h2>\n\n\n\n<p>A <strong>model card<\/strong> is a structured disclosure document for a machine learning model. It was popularized by Mitchell et al. in 2019 as a \u201cnutrition label\u201d for AI models.<\/p>\n\n\n\n<p>Model cards are designed to be:<\/p>\n\n\n\n<ul>\n<li>Short and readable by non-experts.<\/li>\n\n\n\n<li>Standardized across models and teams.<\/li>\n\n\n\n<li>Focused on real-world use and impact.<\/li>\n\n\n\n<li>Useful for developers, product managers, auditors, and affected communities.<\/li>\n<\/ul>\n\n\n\n<p>They complement technical documentation by emphasizing context, ethics, and limitations rather than only architecture and code.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\"><strong>Why Use Model Cards?<\/strong><\/h2>\n\n\n\n<p>Model cards help teams:<\/p>\n\n\n\n<ul>\n<li>Clarify intended and out-of-scope use cases.<\/li>\n\n\n\n<li>Report performance across relevant subgroups.<\/li>\n\n\n\n<li>Document known limitations and failure modes.<\/li>\n\n\n\n<li>Surface ethical and societal considerations.<\/li>\n\n\n\n<li>Support internal reviews and external audits.<\/li>\n\n\n\n<li>Meet regulatory and governance requirements.<\/li>\n<\/ul>\n\n\n\n<p>Without model cards, important information often remains scattered in code, notebooks, or internal chats.<\/p>\n\n\n\n<p><em>Model cards and datasheets are standardized \u201cnutrition labels\u201d for AI models. Model cards document intended use, performance, and limits; datasheets describe dataset provenance, composition, and ethics. Learn AI &amp; ML with HCL GUVI\u2019s<\/em><a href=\"https:\/\/www.guvi.in\/mlp\/artificial-intelligence-and-machine-learning?utm_source=Model-Cards-and-Datasheets-Documenting-AI-Responsibly\" target=\"_blank\" rel=\"noreferrer noopener\"><em> Artificial Intelligence and Machine Learning course<\/em><\/a><em>.&nbsp;<\/em><\/p>\n\n\n\n<h2 class=\"wp-block-heading\"><strong>Typical Sections in a Model Card<\/strong><\/h2>\n\n\n\n<p>While formats vary, most model cards include the following sections.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>1. Model Name and Version<\/strong><\/h3>\n\n\n\n<ul>\n<li>Unique identifier.<\/li>\n\n\n\n<li>Version number and release date.<\/li>\n\n\n\n<li>Link to code, weights, or deployment endpoint.<\/li>\n<\/ul>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>2. Model Type and Architecture<\/strong><\/h3>\n\n\n\n<ul>\n<li>Model family (for example, transformer, gradient boosting).<\/li>\n\n\n\n<li>Architecture details at a high level.<\/li>\n\n\n\n<li>Framework and library versions.<\/li>\n<\/ul>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>3. Intended Use<\/strong><\/h3>\n\n\n\n<ul>\n<li>Primary tasks the model was designed for.<\/li>\n\n\n\n<li>Target users and environments.<\/li>\n\n\n\n<li>Example applications.<\/li>\n<\/ul>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>4. Out-of-Scope Use<\/strong><\/h3>\n\n\n\n<ul>\n<li>Use cases the model should not be used for.<\/li>\n\n\n\n<li>High-risk contexts where the model is not validated.<\/li>\n\n\n\n<li>Examples of misuse to avoid.<\/li>\n<\/ul>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>5. Training Data<\/strong><\/h3>\n\n\n\n<ul>\n<li>Description of training data sources.<\/li>\n\n\n\n<li>Time period, geography, and domains covered.<\/li>\n\n\n\n<li>Preprocessing and cleaning steps.<\/li>\n\n\n\n<li>Reference to the dataset\u2019s datasheet if available.<\/li>\n<\/ul>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>6. Evaluation Data<\/strong><\/h3>\n\n\n\n<ul>\n<li>Description of evaluation and test datasets.<\/li>\n\n\n\n<li>How evaluation data differs from training data.<\/li>\n\n\n\n<li>Any known biases or gaps in evaluation coverage.<\/li>\n<\/ul>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>7. Performance Metrics<\/strong><\/h3>\n\n\n\n<ul>\n<li>Key metrics such as accuracy, precision, recall, F1, <a href=\"https:\/\/www.guvi.in\/blog\/auc-roc-curve-the-real-report-card\/\" target=\"_blank\" rel=\"noreferrer noopener\">AUC<\/a>, and latency.<\/li>\n\n\n\n<li>Confidence intervals where applicable.<\/li>\n\n\n\n<li>Baselines or comparator models.<\/li>\n<\/ul>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>8. Disaggregated Performance<\/strong><\/h3>\n\n\n\n<ul>\n<li>Performance broken down by relevant groups:\n<ul>\n<li>Demographics (for example, gender, age, ethnicity).<\/li>\n\n\n\n<li>Geographies or languages.<\/li>\n\n\n\n<li>Device types or usage contexts.<\/li>\n<\/ul>\n<\/li>\n\n\n\n<li>Identification of groups with notably worse performance.<\/li>\n<\/ul>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>9. Limitations<\/strong><\/h3>\n\n\n\n<ul>\n<li>Known failure modes and edge cases.<\/li>\n\n\n\n<li>Conditions under which performance degrades.<\/li>\n\n\n\n<li>Data or context gaps that affect reliability.<\/li>\n<\/ul>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>10. Ethical Considerations<\/strong><\/h3>\n\n\n\n<ul>\n<li>Potential harms and misuse scenarios.<\/li>\n\n\n\n<li>Bias and fairness observations.<\/li>\n\n\n\n<li>Sensitive use contexts (for example, healthcare, law enforcement).<\/li>\n\n\n\n<li>Mitigation strategies and safeguards.<\/li>\n<\/ul>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>11. Recommendations for Use<\/strong><\/h3>\n\n\n\n<ul>\n<li>Guidance on responsible deployment.<\/li>\n\n\n\n<li>Monitoring and human oversight recommendations.<\/li>\n\n\n\n<li>When to retrain or decommission the model.<\/li>\n<\/ul>\n\n\n\n<figure class=\"wp-block-pullquote\"><blockquote><p><strong>Best Practice <\/strong>Treat the model card as a living document. Update it when the model, data, or usage context changes significantly.<\/p><\/blockquote><\/figure>\n\n\n\n<h2 class=\"wp-block-heading\"><strong>What Are Datasheets for Datasets?<\/strong><\/h2>\n\n\n\n<p><strong>Datasheets for datasets<\/strong> were introduced by Gebru et al. in 2018, inspired by electronics datasheets. They provide structured documentation for the datasets used to train or evaluate AI systems.<br>Datasheets help answer questions such as:<\/p>\n\n\n\n<ul>\n<li>Why was this dataset created?<\/li>\n\n\n\n<li>What data does it contain?<\/li>\n\n\n\n<li>How was it collected and labeled?<\/li>\n\n\n\n<li>Who is represented, and who is missing?<\/li>\n\n\n\n<li>What are the legal and ethical constraints?<\/li>\n\n\n\n<li>How should the dataset be used and maintained?<\/li>\n<\/ul>\n\n\n\n<p>Like model cards, datasheets aim to make hidden assumptions and risks more visible.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\"><strong>Why Use Datasheets?<\/strong><\/h2>\n\n\n\n<p>Datasheets help teams:<\/p>\n\n\n\n<ul>\n<li>Understand dataset provenance and quality.<\/li>\n\n\n\n<li>Identify potential biases and representation gaps.<\/li>\n\n\n\n<li>Clarify appropriate and inappropriate uses.<\/li>\n\n\n\n<li>Support reproducibility and auditing.<\/li>\n\n\n\n<li>Comply with data governance and regulatory requirements.<\/li>\n\n\n\n<li>Communicate dataset limitations to downstream users.<\/li>\n<\/ul>\n\n\n\n<p>Without datasheets, dataset decisions often remain implicit and hard to challenge.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\"><strong>Typical Sections in a Datasheet<\/strong><\/h2>\n\n\n\n<p>Common sections include:<\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>1. Motivation<\/strong><\/h3>\n\n\n\n<ul>\n<li>Why the dataset was created.<\/li>\n\n\n\n<li>What tasks it is intended to support.<\/li>\n\n\n\n<li>Who requested or funded the dataset.<\/li>\n<\/ul>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>2. Composition<\/strong><\/h3>\n\n\n\n<ul>\n<li>What data is included (for example, text, images, tabular records).<\/li>\n\n\n\n<li>Number of instances and features.<\/li>\n\n\n\n<li>Class distribution and label schema.<\/li>\n\n\n\n<li>Demographic or contextual breakdowns where relevant.<\/li>\n<\/ul>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>3. Collection Process<\/strong><\/h3>\n\n\n\n<ul>\n<li>How the data was collected (for example, scraping, surveys, sensors).<\/li>\n\n\n\n<li>Time period and geographic coverage.<\/li>\n\n\n\n<li>Consent procedures and participant information.<\/li>\n\n\n\n<li>Compensation or incentives, if any.<\/li>\n<\/ul>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>4. Preprocessing and Cleaning<\/strong><\/h3>\n\n\n\n<ul>\n<li>Filtering, normalization, and transformation steps.<\/li>\n\n\n\n<li>Handling of missing or invalid data.<\/li>\n\n\n\n<li>Any augmentation or synthetic data generation.<\/li>\n<\/ul>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>5. Labeling<\/strong><\/h3>\n\n\n\n<ul>\n<li>Who labeled the data (for example, crowd workers, experts).<\/li>\n\n\n\n<li>Labeling guidelines and quality control.<\/li>\n\n\n\n<li>Inter-annotator agreement metrics if available.<\/li>\n<\/ul>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>6. Intended Use<\/strong><\/h3>\n\n\n\n<ul>\n<li>Tasks and contexts the dataset is suitable for.<\/li>\n\n\n\n<li>Target users and applications.<\/li>\n\n\n\n<li>Example use cases.<\/li>\n<\/ul>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>7. Not Appropriate For<\/strong><\/h3>\n\n\n\n<ul>\n<li>Tasks and contexts where the dataset should not be used.<\/li>\n\n\n\n<li>High-risk or sensitive applications.<\/li>\n\n\n\n<li>Known misuse scenarios.<\/li>\n<\/ul>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>8. Distribution<\/strong><\/h3>\n\n\n\n<ul>\n<li>How the dataset is shared (for example, public, restricted, internal).<\/li>\n\n\n\n<li>License terms and access conditions.<\/li>\n\n\n\n<li>Privacy and security controls.<\/li>\n<\/ul>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>9. Maintenance<\/strong><\/h3>\n\n\n\n<ul>\n<li>Who maintains the <a href=\"https:\/\/www.guvi.in\/hub\/fake-review-detection-system-using-machine-learning\/understanding-the-dataset\/\" target=\"_blank\" rel=\"noreferrer noopener\">dataset<\/a>.<\/li>\n\n\n\n<li>How updates, corrections, and versioning are handled.<\/li>\n\n\n\n<li>How users can report issues or request changes.<\/li>\n<\/ul>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>10. Ethical and Legal Considerations<\/strong><\/h3>\n\n\n\n<ul>\n<li>Privacy risks and protections.<\/li>\n\n\n\n<li>Bias and representation issues.<\/li>\n\n\n\n<li>Relevant laws and regulations.<\/li>\n\n\n\n<li>Known controversies or concerns.<\/li>\n<\/ul>\n\n\n\n<figure class=\"wp-block-pullquote\"><blockquote><p><strong>Best Practice:<\/strong> Reference the datasheet in the model card instead of duplicating all dataset details. This keeps documentation modular and easier to maintain.<\/p><\/blockquote><\/figure>\n\n\n\n<h2 class=\"wp-block-heading\"><strong>How Model Cards and Datasheets Work Together<\/strong><\/h2>\n\n\n\n<p>Model cards and datasheets are complementary:<\/p>\n\n\n\n<ul>\n<li><strong>Datasheets<\/strong> describe the data: what it is, where it came from, and how it was created.<\/li>\n\n\n\n<li><strong>Model cards<\/strong> describe the model: how it was trained, how it performs, and how it should be used.<\/li>\n<\/ul>\n\n\n\n<p>A typical relationship looks like this:<\/p>\n\n\n\n<p><strong>Documentation Flow Box<\/strong><\/p>\n\n\n\n<pre class=\"wp-block-code\"><code>text\n\nDataset \u2192 Datasheet\n\n&nbsp;&nbsp;&nbsp;\u2193\n\nModel trained on dataset \u2192 Model card\n\n&nbsp;&nbsp;&nbsp;\u2193\n\nSystem or product \u2192 System card (optional)<\/code><\/pre>\n\n\n\n<p>The model card should reference the relevant datasheets, including version information and any train\/test splits.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\"><strong>Example Mapping<\/strong><\/h2>\n\n\n\n<ul>\n<li><strong>Datasheet sections used in model card:<\/strong>\n<ul>\n<li><a href=\"https:\/\/www.guvi.in\/blog\/guide-to-data-exploration\/\" target=\"_blank\" rel=\"noreferrer noopener\">Data <\/a>composition and source.<\/li>\n\n\n\n<li>Collection and preprocessing.<\/li>\n\n\n\n<li>Intended and prohibited uses.<\/li>\n\n\n\n<li>Ethical and legal notes.<\/li>\n<\/ul>\n<\/li>\n\n\n\n<li><strong>Model card sections that go beyond datasheets:<\/strong>\n<ul>\n<li>Model architecture and training procedure.<\/li>\n\n\n\n<li>Performance metrics and disaggregated results.<\/li>\n\n\n\n<li>Model-specific limitations and failure modes.<\/li>\n\n\n\n<li>Deployment and monitoring recommendations.<\/li>\n<\/ul>\n<\/li>\n<\/ul>\n\n\n\n<h2 class=\"wp-block-heading\"><strong>When Are Model Cards and Datasheets Required?<\/strong><\/h2>\n\n\n\n<p>These documents are increasingly expected or required by:<\/p>\n\n\n\n<ul>\n<li>Internal <a href=\"https:\/\/www.guvi.in\/blog\/what-is-artificial-intelligence\/\" target=\"_blank\" rel=\"noreferrer noopener\">AI <\/a>governance and review boards.<\/li>\n\n\n\n<li>Regulatory frameworks such as the EU AI Act for high-risk systems.<\/li>\n\n\n\n<li>Industry standards and certification schemes.<\/li>\n\n\n\n<li>Customer and partner due-diligence processes.<\/li>\n\n\n\n<li>Academic and research publication norms.<\/li>\n<\/ul>\n\n\n\n<p>Under the EU AI Act, technical documentation for high-risk AI must exist before deployment and be retained for years. Model cards and datasheets can form a core part of this documentation.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\"><strong>Common Mistakes to Avoid<\/strong><\/h2>\n\n\n\n<ul>\n<li>Writing model cards only at the end of a project.<\/li>\n\n\n\n<li>Treating them as marketing materials instead of honest disclosures.<\/li>\n\n\n\n<li>Omitting disaggregated performance and subgroup analysis.<\/li>\n\n\n\n<li>Ignoring out-of-scope and high-risk use cases.<\/li>\n\n\n\n<li>Failing to link model cards to dataset datasheets.<\/li>\n\n\n\n<li>Not updating documents when models or data change.<\/li>\n\n\n\n<li>Using vague language instead of concrete metrics and examples.<\/li>\n\n\n\n<li>Writing only for experts, not for broader stakeholders.<\/li>\n\n\n\n<li>Keeping documents internal when external transparency is needed.<\/li>\n\n\n\n<li>Creating documentation that no one reads or maintains.<\/li>\n<\/ul>\n\n\n\n<div style=\"background-color: #099f4e; border: 3px solid #110053; border-radius: 12px; padding: 18px 22px; color: #FFFFFF; font-size: 18px; font-family: Montserrat, Helvetica, sans-serif; line-height: 1.6; box-shadow: 0 4px 12px rgba(0, 0, 0, 0.15); max-width: 750px;\"> \n  <strong style=\"font-size: 22px; color: #FFFFFF;\">\ud83d\udca1 Did You Know?<\/strong> \n  <br \/><br \/> \n   Studies have found that only a small fraction of public model cards document ethical considerations or subgroup performance in detail.\nThat gap makes it harder for users to understand risks and for organizations to meet emerging governance expectations.\n<\/div>\n\n\n\n<p><em>Model cards and datasheets are standardized \u201cnutrition labels\u201d for AI models. Model cards document intended use, performance, and limits; datasheets describe dataset provenance, composition, and ethics. Learn AI &amp; ML with HCL GUVI\u2019s<\/em><a href=\"https:\/\/www.guvi.in\/mlp\/artificial-intelligence-and-machine-learning?utm_source=Model-Cards-and-Datasheets-Documenting-AI-Responsibly\" target=\"_blank\" rel=\"noreferrer noopener\"><em> Artificial Intelligence and Machine Learning course<\/em><\/a><em>.&nbsp;<\/em><\/p>\n\n\n\n<h2 class=\"wp-block-heading\"><strong>What Should You Do Next?<\/strong><\/h2>\n\n\n\n<p>Use this practical checklist:<\/p>\n\n\n\n<ol>\n<li>Identify which models and datasets need documentation.<\/li>\n\n\n\n<li>Adopt or adapt standard model card and datasheet templates.<\/li>\n\n\n\n<li>Involve data scientists, engineers, product, legal, and ethics <a href=\"https:\/\/www.sigmamagic.com\/blogs\/best-practices-for-stakeholder-analysis\/\" target=\"_blank\" rel=\"noreferrer noopener nofollow\">stakeholders<\/a>.<\/li>\n\n\n\n<li>Document intended use, limitations, and performance early.<\/li>\n\n\n\n<li>Include disaggregated metrics and subgroup analysis where relevant.<\/li>\n\n\n\n<li>Reference datasheets from model cards to avoid duplication.<\/li>\n\n\n\n<li>Review and update documents when models or data change.<\/li>\n\n\n\n<li>Integrate documentation into your release and governance process.<\/li>\n\n\n\n<li>Make documents accessible to relevant internal and external audiences.<\/li>\n\n\n\n<li>Treat model cards and datasheets as living artifacts, not one-off reports.<\/li>\n<\/ol>\n\n\n\n<h2 class=\"wp-block-heading\"><strong>Conclusion<\/strong><\/h2>\n\n\n\n<p><strong>Model cards and datasheets<\/strong> provide structured, human-readable documentation for AI models and datasets. They clarify intended use, performance, data provenance, limitations, and ethical considerations in a way that supports responsible development and deployment.<\/p>\n\n\n\n<p>As AI governance matures, these documents are moving from optional best practice to expected or required artifacts. Teams that invest in clear, honest, and maintained documentation will be better positioned to manage risk, earn trust, and comply with emerging regulations.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\"><strong>FAQs<\/strong><\/h2>\n\n\n<div id=\"rank-math-faq\" class=\"rank-math-block\">\n<div class=\"rank-math-list \">\n<div id=\"faq-question-1787736855656\" class=\"rank-math-list-item\">\n<h3 class=\"rank-math-question \"><strong>What is a model card?<\/strong><\/h3>\n<div class=\"rank-math-answer \">\n\n<p>A model card is a standardized document that describes a trained AI model\u2019s intended use, performance, training data, limitations, and ethical considerations.<\/p>\n\n<\/div>\n<\/div>\n<div id=\"faq-question-1787736862758\" class=\"rank-math-list-item\">\n<h3 class=\"rank-math-question \"><strong>What is a datasheet for a dataset?<\/strong><\/h3>\n<div class=\"rank-math-answer \">\n\n<p>A datasheet for a dataset documents the dataset\u2019s motivation, composition, collection process, preprocessing, labeling, distribution, maintenance, and ethical and legal considerations.<\/p>\n\n<\/div>\n<\/div>\n<div id=\"faq-question-1787736871148\" class=\"rank-math-list-item\">\n<h3 class=\"rank-math-question \"><strong>Why are model cards and datasheets important?<\/strong><\/h3>\n<div class=\"rank-math-answer \">\n\n<p>They improve transparency, accountability, and reproducibility. They help stakeholders understand capabilities, limitations, and risks, and support governance and regulatory compliance.<\/p>\n\n<\/div>\n<\/div>\n<div id=\"faq-question-1787736888719\" class=\"rank-math-list-item\">\n<h3 class=\"rank-math-question \"><strong>Who should write model cards and datasheets?<\/strong><\/h3>\n<div class=\"rank-math-answer \">\n\n<p>Ideally, a cross-functional team including data scientists, engineers, product managers, legal or compliance, and ethics or policy representatives.<\/p>\n\n<\/div>\n<\/div>\n<div id=\"faq-question-1787736898114\" class=\"rank-math-list-item\">\n<h3 class=\"rank-math-question \"><strong>When should model cards and datasheets be created?<\/strong><\/h3>\n<div class=\"rank-math-answer \">\n\n<p>Early in the project, and updated whenever the model, data, or usage context changes significantly. They should be treated as living documents.<\/p>\n\n<\/div>\n<\/div>\n<div id=\"faq-question-1787736906336\" class=\"rank-math-list-item\">\n<h3 class=\"rank-math-question \"><strong>Are model cards and datasheets required by law?<\/strong><\/h3>\n<div class=\"rank-math-answer \">\n\n<p>Not universally, but they are increasingly expected or required by governance frameworks and regulations such as the EU AI Act for high-risk AI systems.<\/p>\n\n<\/div>\n<\/div>\n<div id=\"faq-question-1787736915485\" class=\"rank-math-list-item\">\n<h3 class=\"rank-math-question \"><strong>Should model cards be public?<\/strong><\/h3>\n<div class=\"rank-math-answer \">\n\n<p>It depends on the context. Internal model cards support governance and handover. Public model cards increase transparency for users, researchers, and affected communities.<\/p>\n\n<\/div>\n<\/div>\n<div id=\"faq-question-1787736924235\" class=\"rank-math-list-item\">\n<h3 class=\"rank-math-question \"><strong>How do model cards relate to datasheets?<\/strong><\/h3>\n<div class=\"rank-math-answer \">\n\n<p>Model cards describe models; datasheets describe datasets. Model cards should reference relevant datasheets to avoid duplication and provide a complete picture of the AI system.<\/p>\n\n<\/div>\n<\/div>\n<\/div>\n<\/div>","protected":false},"excerpt":{"rendered":"<p>Model cards and datasheets are structured documents that describe AI models and datasets in a clear, standardized way. They help teams communicate capabilities, limitations, and ethical considerations to developers, auditors, and end users. As regulations and governance frameworks mature, these documents are becoming essential for responsible AI deployment. TL;DR Summary Direct Answer Model cards are [&hellip;]<\/p>\n","protected":false},"author":7,"featured_media":135973,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[933],"tags":[],"views":"13","authorinfo":{"name":"HCL GUVI","url":"https:\/\/www.guvi.in\/blog\/author\/guvipr\/"},"thumbnailURL":"https:\/\/www.guvi.in\/blog\/wp-content\/uploads\/2026\/08\/Model-Cards-and-Datasheets-Documenting-AI-Responsibly-300x116.webp","_links":{"self":[{"href":"https:\/\/www.guvi.in\/blog\/wp-json\/wp\/v2\/posts\/135766"}],"collection":[{"href":"https:\/\/www.guvi.in\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.guvi.in\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.guvi.in\/blog\/wp-json\/wp\/v2\/users\/7"}],"replies":[{"embeddable":true,"href":"https:\/\/www.guvi.in\/blog\/wp-json\/wp\/v2\/comments?post=135766"}],"version-history":[{"count":4,"href":"https:\/\/www.guvi.in\/blog\/wp-json\/wp\/v2\/posts\/135766\/revisions"}],"predecessor-version":[{"id":137228,"href":"https:\/\/www.guvi.in\/blog\/wp-json\/wp\/v2\/posts\/135766\/revisions\/137228"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.guvi.in\/blog\/wp-json\/wp\/v2\/media\/135973"}],"wp:attachment":[{"href":"https:\/\/www.guvi.in\/blog\/wp-json\/wp\/v2\/media?parent=135766"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.guvi.in\/blog\/wp-json\/wp\/v2\/categories?post=135766"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.guvi.in\/blog\/wp-json\/wp\/v2\/tags?post=135766"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}