--- title: "Model Cards for AI/ML SaMD · AI SaMD Playbook" description: "Template, twelve-section checklist and worked examples for AI/ML SaMD model cards that hold up under FDA, MHRA, HC and EU notified body review." lang: en json-ld: | [ { "@context": "https://schema.org", "@graph": [ { "@type": "WebSite", "@id": "https://ai-samd.com/#website", "url": "https://ai-samd.com/", "name": "AI SaMD Playbook", "description": "Working reference for regulating AI/ML Software as a Medical Device: FDA, EU AI Act, MHRA, Health Canada and NMPA guidance, model risks, and cybersecurity for RA/QA teams.", "inLanguage": "en" }, { "@type": "Organization", "name": "Blue Goat Cyber", "url": "https://bluegoatcyber.com" } ] }, { "@context": "https://schema.org", "@type": "BreadcrumbList", "itemListElement": [ { "@type": "ListItem", "position": 1, "name": "Home", "item": "https://ai-samd.lovable.app/" }, { "@type": "ListItem", "position": 2, "name": "Model Cards", "item": "https://ai-samd.lovable.app/model-cards" } ] } ] --- [AI SaMD Playbook ](/) Sections[Crosswalk](/#crosswalk)Tools Search Ctrl K[Talk to us →](https://go.bluegoatcyber.com/meetings/blue-goat-cyber/discovery-session) [← Back to home](/) Starter guide · v1 # Model cards _that hold up_ in a regulatory review. A working template, twelve-section checklist, and worked examples for assembling a model card that answers what FDA, MHRA, Health Canada, and EU notified bodies actually ask. Mapped to the guiding documents already cited on this site. Download checklist (.md)[Jump to sections ↓](#sections) Built on - FDA Transparency Guiding Principles - FDA / Health Canada / MHRA, Good Machine Learning Practice - FDA PCCP final guidance · Aug 2025 - EU AI Act Art. 13 & 113 · MDCG 2025-6 Six anchor sections ## The structure reviewers expect to find. Each section pairs the regulator's question with a working example of the answer. Use the downloadable checklist for the full twelve-section version. §01 ### Intended use & indications Anchor every downstream claim. Reviewers compare every metric and limitation against the words you put here. [Source · FDA, Transparency for ML-enabled devices ↗](https://www.fda.gov/medical-devices/software-medical-device-samd/transparency-machine-learning-enabled-medical-devices-guiding-principles) Required fields - Clinical task in plain language - Target patient population (age, sex, comorbidities) - Care setting and user profile - Explicit out-of-scope uses & contraindications Worked example Task Triage of non-contrast head CT for suspected intracranial haemorrhage Population Adults ≥18 presenting to ED with suspected acute stroke User Board-certified radiologist; not for use without expert review Out of scope Paediatric, post-operative, or contrast-enhanced studies §02 ### Training data provenance Reviewers want to know who is represented, who isn't, and why. Gaps disclosed up front are mitigations; gaps discovered later are findings. [Source · FDA, GMLP guiding principles ↗](https://www.fda.gov/medical-devices/software-medical-device-samd/good-machine-learning-practice-medical-device-development-guiding-principles) Required fields - Sources, institutions, geographies, years - Sample size + class balance - Demographic + device distribution - Inclusion / exclusion criteria - Labelling protocol + inter-rater agreement - Known representational gaps Worked example Sources 4 US academic centres + 1 EU teaching hospital, 2018–2023 N 12,840 studies; 18.4% positive for ICH Scanners GE, Siemens, Canon, 64-slice and above Gap Under-representation of patients <30 and non-contrast scanners <16-slice §03 ### Performance, overall and by subgroup Headline metrics are not enough. Stratified results are now an explicit expectation in FDA, MHRA and Health Canada review. [Source · FDA, GMLP guiding principles ↗](https://www.fda.gov/medical-devices/software-medical-device-samd/good-machine-learning-practice-medical-device-development-guiding-principles) Required fields - Independent test set definition (site / patient / time split) - Primary metrics with 95% CIs - Subgroup table: sex, age band, race/ethnicity, device, site - Calibration plot or ECE - External validation cohort Worked example Sensitivity 94.1% (92.7–95.3) Specificity 89.6% (88.4–90.7) Sens. (female ≥65) 91.0% (87.8–93.6), flagged for monitoring External cohort n=2,104, EU site, sens. 92.4% / spec. 88.1% §04 ### Limitations & failure modes Documenting where the model breaks is a transparency obligation under both the FDA guiding principles and EU AI Act Article 13. [Source · EU AI Act, consolidated text (Article 13) ↗](https://eur-lex.europa.eu/eli/reg/2024/1689/oj) Required fields - Documented failure modes - Populations / settings where performance is degraded - Open bugs and CAPAs disclosed to deployers - Behaviour on out-of-distribution inputs Worked example Failure mode Motion-degraded studies → confidence < 0.6 returned with abstention Degraded setting Non-contrast scanners <16-slice not validated; device blocks inference OOD Paediatric input → input-validation error, no score returned §05 ### Cybersecurity posture AI inherits every classical software threat and adds adversarial inputs, prompt injection, and model extraction. Both FDA and EU MDR Annex I §17 expect explicit handling. [Source · FDA, Premarket cybersecurity guidance (2023) ↗](https://www.fda.gov/regulatory-information/search-fda-guidance-documents/cybersecurity-medical-devices-quality-system-considerations-and-content-premarket-submissions) Required fields - SBOM including model weights + inference runtime - Threat model + risk register reference - Adversarial-robustness testing for the modality in scope - Coordinated vulnerability disclosure contact Worked example SBOM SPDX 2.3 generated per release; weights pinned by SHA-256 Adversarial FGSM + PGD evaluated; degradation < 3% at ε=2/255 CVD security@vendor.example, 90-day disclosure window §06 ### Lifecycle & change control (PCCP) FDA's August 2025 final PCCP guidance defines what can change without a new submission. Anything not in the PCCP triggers a new authorisation. [Source · FDA, PCCP final guidance (Aug 2025) ↗](https://www.fda.gov/regulatory-information/search-fda-guidance-documents/predetermined-change-control-plans-artificial-intelligence-enabled-device-software-functions) Required fields - PCCP on file? Reference document ID - Modification protocol: what can change, how, with what limits - Performance monitoring metrics + alert thresholds - Drift detection method - Rollback path Worked example PCCP scope Re-training on +20% data per quarter; threshold tuning ±0.05 Monitoring Weekly sensitivity by site; alert at −2σ from baseline Rollback Previous model retained for 24 months; rollback ≤ 4h Take it with you ### The full twelve-section checklist, as Markdown. Drop it in your DMS, paste it into your QMS, or hand it to the model team. Adapted from the same primary sources cited across the rest of this site. Download checklist (.md) © 2026 AI SaMD Playbook · Sponsored by Blue Goat Cyber [Crosswalk](/crosswalk)[Colophon](/colophon) Editorial reference only. Not legal or regulatory advice.