GenAI-Gurus/awesome-eu-ai-act
GenAI-Gurus/awesome-eu-ai-act has 135 GitHub stars and 85 forks. GitHot measured +1 stars in 24 hours. Last checked 2026-11-16 18:54 UTC. Last push 2026-11-08.
Measured star growth
- in 24h
- +1
- in 7 days
- Collecting history
- in 15 days
- Collecting history
- in 1 month
- Collecting history
Net star changes over overlapping periods, not daily totals. Missing measurements are left blank. Stars measure attention, not software quality.
About
Curated tools, official sources, OSS, templates, and guides for EU AI Act compliance.
AIREADME.md
OriginalAwesome EU AI Act 
EU regulation governing the development and deployment of artificial intelligence systems, establishing a risk-based compliance framework with phased implementation timelines.
Contents
- Official EU Sources
- Compliance Tools & Platforms
- Open-Source Projects
- Legal Analysis & Guides
- Standards & Frameworks
- Templates & Checklists
- Training & Certifications
- News & Newsletters
- Community & Events
- AI Governance & Ethics Tools
- Regulatory Sandboxes
- Related EU Regulations
Official EU Sources
Regulation Text
- EU AI Act Full Text (EUR-Lex) - Regulation (EU) 2024/1689 as published in the Official Journal on 12 July 2024.
- EU AI Act Consolidated Text (EUR-Lex) - Consolidated text incorporating the Digital Omnibus on AI amendments.
- Digital Omnibus on AI (EUR-Lex) - Regulation (EU) 2026/1744 amending the AI Act, in force since 27 July 2026.
- AI Act PDF Version (EUR-Lex) - Official PDF version of the AI Act.
- EUR-Lex Plain-Language Summary - Accessible summary of key provisions and the risk-based approach.
- Original 2021 Proposal (COM(2021) 206) - The European Commission's original legislative proposal for historical context.
European Commission
- AI Act - Shaping Europe's Digital Future - Commission's main AI Act landing page with implementation timeline and guidelines.
- European Approach to Artificial Intelligence - Overarching EU AI strategy and AI Continent Action Plan.
- AI Omnibus Implementation Update - Commission summary of the adopted changes and revised application dates.
- AI Act Standardisation - Official page on CEN/CENELEC harmonized standards in ten key areas.
- Supporting Implementation with Clear Guidelines - Published and upcoming guidelines on prohibited practices, high-risk classification, and transparency.
- Guidelines on Prohibited AI Practices - Commission's interpretation of Article 5 prohibitions with practical examples (4 February 2025).
- Guidelines on the Definition of an AI System - Commission breakdown of the seven cumulative elements of the Article 3(1) AI system definition.
- Guidelines on the Scope of Obligations for Providers of GPAI Models - Commission interpretation of GPAI definition, the 10²³ FLOP threshold, and downstream-provider duties.
- Article 50 Transparency Guidelines - Commission guidance on Article 50 duties and exceptions for providers and deployers.
- Code of Practice on Marking and Labelling AI-Generated Content - Final voluntary Article 50 code for marking, detection, and labelling workflows.
- AI Literacy Q&A - Official FAQ on AI literacy obligations under Article 4.
- AI Pact - Voluntary initiative enabling early compliance before legal deadlines.
EU AI Office
- European AI Office - Coordinates AI Act implementation and enforces GPAI rules.
- AI Act Single Information Platform - Interactive compliance tools, AI Act Explorer, Compliance Checker, and FAQs.
- EU AI Act Compliance Checker (EC) - Official beta tool for determining obligations.
GPAI Code of Practice
- General-Purpose AI Code of Practice - Voluntary tool for GPAI providers to demonstrate AI Act compliance on transparency, copyright, and safety.
- GPAI Code of Practice (European Commission) - Official Commission page with the GPAI Code of Practice signatory list, signature process, and Q&A.
European Parliament & Governance
- EU AI Act Explainer (European Parliament) - Parliament's guide covering risk levels, prohibitions, and timeline.
- AI Act Simplification Vote (European Parliament) - Parliament's June 2026 vote on the AI Omnibus.
- Digital Omnibus on AI Procedure File - Legislative tracker for the Digital Omnibus on AI file.
- European AI Board - Member State representatives advising on consistent implementation.
- Ethics Guidelines for Trustworthy AI (HLEG) - Landmark 2019 seven-requirement framework that shaped the AI Act.
Unofficial Authoritative References
- EU Artificial Intelligence Act (Future of Life Institute) - Most comprehensive independent resource; 150,000+ monthly users. Includes AI Act Explorer and Compliance Checker.
- AI Act Explorer (FLI) - Interactive browsing by title, chapter, article, and recital.
- EU AI Act Compliance Checker (FLI) - Free tool to determine whether an AI system falls under AI Act obligations.
- AI Law Radar - Primary-sourced tracker for AI regulation obligations and deadlines, including EU AI Act updates.
- EU AI Act Dates After the Digital Omnibus (Meilynx) - Each obligation, its article, and its application date as amended by Regulation (EU) 2026/1744.
Compliance Tools & Platforms
Enterprise AI Governance Platforms
- Holistic AI - Enterprise AI governance with lifecycle oversight, bias testing, and EU AI Act readiness assessment.
- Credo AI - AI governance, risk, and compliance platform with pre-built policy packs for EU AI Act, NIST AI RMF, ISO 42001, and SOC 2.
- OneTrust AI Governance - AI use case intake/approval workflows, unified asset inventory, and real-time monitoring.
- IBM watsonx.governance - Enterprise AI governance for monitoring fairness, quality, explainability, and drift in ML models.
- ValidMind - AI governance platform with centralized governance, policies, workflows, and responsible AI controls.
- Lumenova AI - Responsible AI governance with automated fairness assessments and governance workflows for non-technical staff.
- Modulos - Swiss-based AI governance platform connecting frameworks, requirements, controls, and evidence into a governance graph.
- Dataiku EU AI Act Readiness - Governance workflow blueprints and compliance management.
- Fiddler AI - Full-stack AI observability with model monitoring, SHAP-based explainability, and fairness checks.
EU AI Act-Specific Tools
- Vanta EU AI Act - Compliance automation with risk classification, evidence reuse from NIST AI RMF/ISO 42001, and continuous monitoring.
- FairNow - Centralized AI inventory, automated bias audits, and a 2-minute AI Compliance Checker.
- Kertos - EU-focused platform supporting EU AI Act, GDPR, and ISO 27001 with AI-driven automation.
- Hyperproof - Compliance operations with automation for evidence collection and control mapping.
- ComplyACT AI - Auto-classifies AI systems, generates audit-ready Annex IV documentation, and provides continuous monitoring.
- SetAIComply - Free client-side AI Act risk-tier checker in all 24 EU languages; embeddable.
- ComplyLayer - AI inventory, risk classification, role-scoped documents and Article 4 acknowledgement records.
- AI Disclosure Kit - Generates Article 50 disclosure templates for AI transparency notices.
- AI Interaction Transparency Snippets - Source-traced examples for AI-chat and synthetic-content notices.
- The AI Act Clock - AI Act obligation and deadline questionnaire with an open JSON ruleset.
- EU AI Act Check - German-language browser tool for preliminary AI Act risk classification and obligations.
- Scanara - Scans code and docs, classifies AI system risk, and generates EU AI Act compliance dossiers.
- Retrace - Open-source scanner that reports AI chatbots and AI-notice wording on a website (Art. 50 evidence).
Open-Source Projects
EU AI Act Compliance Platforms
- ai-act-conformity-pack - Generates an Annex IV technical-documentation skeleton (Articles 11-15) from a YAML model card. Python CLI, MIT.
- ai-act-skills - Multi-platform agent skills for EU AI Act compliance workflows anchored in ISO/IEC 42001 and ISO/IEC 27090.
- VerifyWise - Complete AI governance and LLM evals platform with support for EU AI Act, ISO 42001, NIST AI RMF, and 20+ frameworks (~247 stars).
- EuConform - Risk classification, bias detection via CrowS-Pairs, and Annex IV-compliant PDF report generation, 100% offline (~107 stars).
- EU AI Act MCP Server - MCP server with compliance tools including risk classification, prohibited practice checks, and transparency disclosures.
- MCP EU AI Act Scanner - Scans codebases for EU AI Act and GDPR compliance gaps and generates auditor-ready Annex IV evidence packages.
- Compl-AI - Compliance-centered LLM evaluation framework with technical interpretation of the AI Act and benchmarking suite covering six core principles.
- @eucomplyhub/mcp-eu-ai-act - MCP server for EU AI Act risk classification and Annex III analysis.
- PALO Framework - Local-first EU AI Act risk tiering, FRIA, lifecycle controls, and evidence workflows.
- GOPAL - OPA/Rego policy library with 29 tested checks across 19 EU AI Act articles.
Reference Implementations
- PRML / falsify - Open specification for pre-registering ML evaluation claims with tamper-evident hashes.
- Practical AI Act - Implementation of a high-risk AI system per Chapter III using an open-source ML pipeline (Bavarian AI Act Accelerator).
- AI Act Engineering - Curated reference list for engineering practices ensuring AI systems comply with AI Act regulations.
- AI Act Technical Documentation Assessment Tools - Research replication package for using AI to draft Annex IV-compliant technical documentation.
- AIR Blackbox - Open-source trust infrastructure with 39 EU AI Act compliance checks, decision traceability, and audit chains across 11 PyPI packages.
Risk Assessment & Classification
- AI Assessment Tool (AI4Belgium) - Open-source interactive tool based on ALTAI recommendations for assessing AI trustworthiness.
- AI Act Implementation Tool (Algorithm Audit) - Risk classification of algorithmic systems using simplified questionnaires.
- Algoritmekader (Dutch Government) - Netherlands' open-source Algorithm Framework for lawful and ethical government AI use.
- Regula - CLI for EU AI Act risk scanning, conformity evidence packs, and CrowS-Pairs bias evaluation.
- EU AI Act Classifier - Deterministic risk-tier classifier with pinpoint citations, obligation tracking, and review gates.
- AI Act Companion - Local-first EU AI Act risk classifier with cited results and document templates.
- High-Risk Classifier (EU AI Regulation Decoded) - Article 6 / Annex III decision aid mapping an AI system to its risk tier and expected evidence.
- EU AI Act Risk Checker (Crux Digits) - Browser-based risk-tier triage with per-article citations, updated for the 2026 Digital Omnibus.
Educational & Informational
- EU AI Act Interactive Guide - Comprehensive interactive guide for startups and SMEs with compliance checker and risk assessment.
- Hugging Face EU AI Act Guide for OSS Developers - Official guide covering obligations for GPAI models, copyright compliance, and documentation requirements.
AI Agent Governance
- Microsoft Agent Governance Toolkit - Seven-package, MIT-licensed runtime governance system for autonomous AI agents covering all 10/10 OWASP Agentic Top 10 (2026) with EU AI Act, NIST AI RMF, HIPAA, and SOC 2 mappings. Enforces policy at <0.1ms p99 latency across LangChain, CrewAI, AutoGen, OpenAI Agents, Google ADK, and 8+ agent frameworks. Python, TypeScript, .NET, Rust, and Go (~900+ stars).
- agent-security-harness - Adversarial testing framework for autonomous agents with EU AI Act crosswalks.
- Vaara - Python runtime evidence layer for AI agents: policy-gated tool calls, hash-chained tamper-evident audit trails with external time anchoring, and independently verifiable execution receipts, aligned with EU AI Act Article 14 (human oversight) and Article 12 (record-keeping). AGPL-3.0-or-later.
- EATF - Appraisal profiles, verifiers, and test vectors for AI agent action evidence.
- Bifrost - Open-source Go gateway for LLM and MCP traffic with routing, request logging, and usage controls.
- Mandare - New open-source gateway: spend limits and a tamper-evident ledger for AI agent LLM calls.
- Tracekit - Gates agent tool calls (allow, deny, ask a person) and signs each into an offline-verifiable log.
- solvi - Python decision runtime: rule checks, human escalation, hash-chained replayable audit trail.
AGT Implementation References
- AGT EU AI Act Compliance Checklist - Maps Agent Governance Toolkit controls to EU AI Act requirements.
- AGT NIST AI RMF Alignment - Maps Agent Governance Toolkit controls to the NIST AI Risk Management Framework.
- AGT SOC 2 Type II Mapping - Self-assessment mapping Agent Governance Toolkit controls to SOC 2 criteria.
- AGT Known Limitations - Documents Agent Governance Toolkit design boundaries and compliance gaps.
Curated Lists
- Awesome AI Regulation - Broader AI regulation ecosystem mapping including guidelines, principles, and standards (~800 stars).
- Awesome Responsible AI - Curated list of responsible, trustworthy, and human-centered AI resources.
Legal Analysis & Guides
Major Law Firm Guides
- EU AI Act Guide (Bird & Bird) - Comprehensive PDF guide covering scope, obligations, and compliance requirements.
- EU AI Act Handbook (White & Case) - Actionable compliance guidance over legal theory.
- EU AI Act Guide for In-House Lawyers (Hunton Andrews Kurth) - Practical roadmap with decision trees for risk classification.
- AI Act: 10 Key Points (Linklaters) - Concise ten-point summary of key provisions and obligations.
- GPAI Guidelines Analysis (DLA Piper) - Analysis of the Commission's guidelines on GPAI obligations.
- EU AI Act Q&A (CMS) - Q&A format guide explaining what businesses need to know.
- AI Act Hub (Taylor Wessing) - Hub with analyses, expert contacts, and obligation summary tables.
- AI Watch: Global Regulatory Tracker (White & Case) - AI Act plus 30+ jurisdictions covered.
Think Tanks & Institutions
- EU AI Act Expert Explainer (Ada Lovelace Institute) - Detailed description of significance, scope, and main provisions.
- Implementing the European AI Act (Future of Life Institute) - FLI's main project page with position papers and policy recommendations.
- EU's AI Power Play (Carnegie Endowment) - Analysis of EU's balancing act between enforcement and innovation.
- Limited Global Impact of the EU AI Act (Brookings) - Analysis of the Brussels effect and its limitations.
IAPP Resources
- EU AI Act Topic Page (IAPP) - Comprehensive aggregation of all IAPP EU AI Act resources.
- EU AI Act Implementation Timeline (IAPP) - Timeline infographic showing key milestones.
Academic Research
- Foundation Models under the EU AI Act (Stanford CRFM) - Analysis of 25 requirements for general-purpose AI models.
- Regulating AI in Europe: Four Problems and Four Solutions (SSRN) - Influential paper analyzing structural flaws and recommending solutions.
- Interplay Between the AI Act and EU Digital Framework (European Parliament) - Comprehensive study examining AI Act interaction with GDPR, DSA, DMA, Data Act, CRA, and NIS2.
- Analyzing the EU AI Act (Stanford HAI) - Expert panel analysis examining benefits, pitfalls, and areas for improvement.
Standards & Frameworks
ISO/IEC Standards
- ISO/IEC 42001:2023 - AI Management System - World's first certifiable AI management system standard (AIMS).
- ISO 42001 Visual Library - Visual learning library for ISO/IEC 42001:2023 covering clauses, Annex A controls, and the PDCA cycle through infographics.
- ISO/IEC 23894:2023 - AI Risk Management - AI-specific risk management guidance adapting ISO 31000 to AI system lifecycles.
- ISO/IEC 38507:2022 - Governance of AI - Governance guidance for organizational AI use and leadership accountability.
- ISO/IEC 42005:2025 - AI Impact Assessment - Framework for understanding AI system impacts on individuals, groups, and society.
- ISO AI Standards Hub - Central page listing all published and in-development ISO AI standards.
NIST AI Risk Management Framework
- NIST AI RMF - NIST's central page for AI standards including the AI RMF.
- NIST AI RMF Crosswalks - Official mappings to ISO/IEC 23894 and the EU AI Act.
- ISO 42001 & NIST AI RMF for EU AI Act Compliance (CSA) - Cloud Security Alliance mapping specific AI Act articles to both frameworks.
CEN/CENELEC Standardization (JTC 21)
- EN 18286:2026 Official Journal Reference - Quality management standard with a presumption of conformity for covered Article 17 requirements.
- CEN-CENELEC JTC 21 - Official body developing European harmonized standards for the AI Act.
- Standard Setting Overview (FLI) - Comprehensive overview of the standard-setting process, JTC 21 structure, and timeline.
- CEN-CENELEC AI Standards - Trustworthiness, risk management, and conformity assessment standards.
ENISA Cybersecurity
- Cybersecurity of AI and Standardisation (ENISA) - Assessment of AI cybersecurity standards with gap analysis.
- Multilayer Framework for AI Cybersecurity (ENISA) - Three-layer framework for securing AI systems.
Data Protection Authorities
- EDPB Opinion 28/2024 on AI Models and GDPR - EDPB position on AI model anonymity, legitimate interest as legal basis, and unlawful training data.
OECD & International
- OECD AI Principles - First intergovernmental AI standard (2019, updated 2024), 47 adherents.
- OECD.AI Policy Observatory - Interactive platform with 900+ national AI policies and tools for building trustworthy AI.
IEEE Standards
- IEEE 7000 Series Standards - Standards covering Transparency (P7001), Data Privacy (P7002), Algorithmic Bias (P7003), and Wellbeing Metrics (P7010).
UK ICO AI Guidance
- Guidance on AI and Data Protection (ICO) - Applying UK GDPR principles to AI systems.
- Explaining Decisions Made with AI (ICO & Alan Turing Institute) - Practical guidance for explaining AI-assisted decisions.
Templates & Checklists
Practical templates, checklists, and assessment materials for AI Act readiness, documentation, FRIAs, and conformity workflows.
Compliance Checklists
- EU AI Act Compliance Matrix (IAPP) - Downloadable charts mapping requirements by operator type across risk categories.
- EU AI Act Compliance Checklist (DPO Europe) - Step-by-step checklist organized by risk category.
- Essential Checklist for EU AI Act Compliance (OneTrust) - Downloadable with scope insights and compliance foundations.
- AI Act Compliance Checklist (Vanta) - Links AI Act compliance to ISO 42001 certification.
- EU AI Act Risk Classification Decision Record - Free editable XLSX for documenting Article 5, Article 6, and Annex III classification evidence.
- EU AI Act 90-Day Implementation Playbook (Secure Privacy) - Structured sprint covering AI system inventory, risk classification, impact assessment, and governance.
- EU AI Regulation Decoded - Open dataset linking EU AI Act obligations to implementation evidence.
- AI Acceptable Use Policy Template (NexoDocs) - Free editable Word policy for staff use of generative AI, supporting Article 4 AI literacy.
FRIA Templates (Fundamental Rights Impact Assessment)
- Guide to FRIAs (Danish Institute for Human Rights) - Practical guide with downloadable FRIA template questionnaire for deployers.
- FRIA Guide (ECNL & Danish Institute) - Five-phase roadmap with downloadable questionnaire.
- FRIA Template (VerifyWise) - Template with impact indicators and risk assessment matrices.
- Free FRIA Template (KLA Digital) - Free template with practical examples for financial services, healthcare, and HR.
Conformity Assessment
- Conformity Assessment Step-by-Step Guide (Future of Privacy Forum) - Detailed walkthrough of the conformity assessment process for high-risk AI systems.
- AI Act Conformity Tool (European DIGITAL SME Alliance) - Interactive tool for SMEs to assess risk level and receive a conformity report.
Technical Documentation (Annex IV)
- Annex IV: Technical Documentation Requirements (FLI) - Official text specifying what must be included.
- Article 11: Technical Documentation (FLI) - Requirements including simplified forms for SMEs.
- Article 17: Quality Management System (FLI) - Full legal text of QMS requirements with 13 mandatory elements.
- Public Summary of Training Content Template (EC) - Mandatory Article 53(1)(d) template for GPAI providers to publicly summarise training data.
Timeline & Governance Infographics
- EU AI Act Implementation Timeline (FPF) - Detailed visual timeline infographic (2025 update).
- EU AI Act Governance Architecture (FPF) - Visual framework mapping the AI Act's institutional architecture.
- EU AI Model Contractual Clauses (European Commission) - Model contractual clauses for AI procurement aligned with the AI Act.
Training & Certifications
Online Courses
- AI and EU Law: Navigating Regulatory Compliance (Coursera) - Covers AI's role in legal processes and compliance with GDPR and AI Act.
- AIGP Certification Masterclass (Udemy) - 19+ hours covering all four AIGP domains aligned with the 2026 Body of Knowledge.
- 40+ EU AI Act Courses (Class Central) - Aggregator listing courses across platforms covering compliance, risk assessment, and implementation.
- Oxford AI Ethics, Regulation and Compliance Programme - Comprehensive programme with Oxford certificate.
- AI Literacy Programs in Europe (FLI) - Curated directory of training programmes supporting Article 4 AI literacy obligations.
- EU AI Act AI Literacy Training Guide (ai-courses.ie) - Ireland-focused guide to Article 4 from EUR-Lex and Commission sources, with courses compared.
Professional Certifications
- AIGP: AI Governance Professional (IAPP) - Premier global certification for AI governance professionals; 100-question exam.
- AI Act Trained Professional / AIActTPro (Cyber Risk GmbH) - Distance learning certification covering all AI Act fundamentals including CE marking.
- A4Q AI Compliance Certification (Learning Tree) - Certified Professional for AI Compliance validating risk classification and audit readiness.
- ForHumanity Certified Auditor - Certification for independent AI system auditing covering 16 core governance pillars.
Enterprise Training
- EU AI Act Mandatory Training (QA) - 40-minute AI literacy training to comply with Article 4.
- EU AI Act Compliance Training (DPO Europe) - Four-week practical training with AIGP-certified trainers.
- RansomLeak EU AI Act Training - Interactive exercises on AI literacy, risk tiers, prohibited practices, and deployer duties.
News & Newsletters
Newsletters
- The EU AI Act Newsletter (Risto Uuk / FLI) - Biweekly; tens of thousands of subscribers; the most detailed AI Act implementation updates available.
- AI Weekly - AI policy coverage, including EU AI Act developments, for 53,000+ professionals.
- Luiza's Newsletter (Luiza Jarovsky, PhD) - 93,000+ subscribers covering AI legal and ethical challenges and rights.
- The AI Policy Newsletter (Alisar Mustafa) - Global AI policy including EU AI Act, US state laws, and international developments.
- Thoughts on AI Policy (MIT & Harvard) - Academic newsletter on AI accountability, transparency, and regulation.
Websites & Blogs
- EU AI Act Updates & Compliance (Cyber Risk GmbH) - Compliance-oriented resource with updates, training materials, and implementation timelines.
- Audaria - French-language coverage of the EU AI Act, ISO/IEC 42001, and AI governance for SMEs.
Podcasts
- The EU AI Act Newsletter Podcast (Spotify) - Audio narrations of the EU AI Act Newsletter by Risto Uuk and FLI.
- The AI Policy Podcast (CSIS) - Biweekly deep dives on AI policy, regulation, and geopolitics.
- RegulatingAI Podcast - Weekly conversations with global leaders on AI governance.
Community & Events
Major Conferences
- IAPP AI Governance Global Europe 2026 (Dublin) - Premier European AI governance conference.
- IAPP Global Summit 2026 - World's largest annual gathering of digital responsibility professionals.
- AI for Good Global Summit 2026 (ITU, Geneva) - UN-organized summit with inaugural Global Dialogue on AI Governance.
Research Organizations
- Centre for the Governance of AI (GovAI, Oxford) - Independent non-profit conducting AI governance research.
- Responsible AI Institute - Offers maturity assessments and RAISE training aligned with NIST, ISO 42001, and EU AI Act.
- IAPP AI Governance Center - Content, networking, training, and certification hub.
AI Governance & Ethics Tools
Explainability Libraries
- SHAP - Game-theoretic ML explainability using Shapley values; the most widely used explainability library (~23,000 stars).
- LIME - Model-agnostic library explaining individual predictions via local surrogate models.
- InterpretML (Microsoft) - Trains interpretable glassbox models (EBM) and explains blackbox systems (~6,000 stars).
- Captum (Meta/PyTorch) - Model interpretability with Integrated Gradients, SHAP, LIME, and saliency maps for PyTorch.
- Alibi (Seldon) - Python library for model inspection: SHAP, ALE, counterfactual explanations, and anchors.
- AI Explainability 360 (IBM) - Comprehensive toolkit for interpretability across tabular, text, image, and time series data (~1,500 stars).
Fairness Toolkits
- AI Fairness 360 (IBM/Linux Foundation) - 70+ fairness metrics and 10+ bias mitigation algorithms (~2,500 stars).
- Fairlearn (Microsoft) - Assessing and improving fairness with group fairness metrics and mitigation algorithms (~1,900 stars).
- Aequitas (University of Chicago) - Bias and fairness audit toolkit across multiple population sub-groups.
- What-If Tool (Google) - Visual, interactive model analysis without code.
- Responsible AI Toolbox (Microsoft) - Unified dashboard combining Fairlearn, InterpretML, Error Analysis, and Counterfactual Analysis (~1,300 stars).
Model Documentation
- Model Card Toolkit (Google/TensorFlow) - Automates generation of Model Cards for transparency reporting (~400 stars).
- Hugging Face Model Cards - Standardized model card format for the Hugging Face Hub.
- AI FactSheets (IBM Research) - Methodology for documenting AI models and services with FactSheets.
- Deon (DrivenData) - Ethics checklist tool adding ethical checkpoints to ML workflows.
ML Monitoring & Observability
- Evidently AI - Open-source platform for evaluating, testing, and monitoring ML/LLM systems with data drift and quality reports.
- Arize AI - Unified AI observability: experiment tracking, production monitoring, and LLM observability.
- WhyLabs - Privacy-preserving AI observability with real-time anomaly detection.
- NannyML - Python library specializing in estimating model performance before ground truth arrives.
Robustness & Adversarial Security
- Adversarial Robustness Toolbox (IBM/Linux Foundation) - Comprehensive framework for evaluating ML security with attacks, defenses, and metrics.
- CleverHans - Library for creating adversarial attacks and benchmarking model robustness.
- Counterfit (Microsoft) - Command-line tool for orchestrating adversarial attacks and testing ML models.
Privacy-Preserving ML
- TensorFlow Privacy - Training ML models with differential privacy.
- PySyft (OpenMined) - Secure and private deep learning supporting federated learning and secure computation.
- OpenDP - Library for privacy-aware computations with differential privacy.
Regulatory Sandboxes
EU AI Act Provisions
- Article 57: AI Regulatory Sandboxes (EUR-Lex) - At least one national AI regulatory sandbox per Member State, operational by 2 August 2027.
- Draft Implementing Act for AI Sandboxes (EC) - Commission consultation on sandbox establishment and operation.
Member State Programs
- EU Member State Sandbox Overview (FLI) - Regularly updated overview of how each Member State is designing and implementing sandboxes.
- AESIA Official Guides (Spain) - Spain's 16 non-binding guidance documents developed through the first EU sandbox pilot; currently the closest thing to a compliance playbook.
Analysis
- Regulatory Sandboxes as AI Governance Tools (FPF) - Analysis of approximately 30 jurisdictions with AI sandboxes.
- Operationalising AI Sandboxes (Cambridge) - Research on capacity, coordination, and provider participation in AI Act sandboxes.
Related EU Regulations
The AI Act does not exist in isolation. These regulations intersect with it and create overlapping obligations.
- GDPR (Regulation (EU) 2016/679) - AI Act data governance requirements (Article 10) must comply with GDPR rules on lawful processing.
- Digital Services Act (Regulation (EU) 2022/2065) - VLOPs using algorithmic recommendation systems face both DSA and AI Act transparency obligations.
- Digital Markets Act (Regulation (EU) 2022/1925) - Gatekeeper platforms deploying AI systems face both DMA interoperability and AI Act high-risk requirements.
- Data Act (Regulation (EU) 2023/2854) - Applicable from September 2025; intersects with AI Act on data access rights for training data.
- Data Governance Act (Regulation (EU) 2022/868) - Enables data sharing that supports AI Act-compliant AI development.
- Product Liability Directive (Directive (EU) 2024/2853) - Explicitly covers software and AI systems; transposition deadline 9 December 2026.
- Cyber Resilience Act (Regulation (EU) 2024/2847) - AI systems that are products with digital elements must meet both CRA and AI Act requirements.
- NIS2 Directive (Directive (EU) 2022/2555) - Entities using AI in critical sectors face both NIS2 incident reporting and AI Act obligations.
- European Health Data Space (Regulation (EU) 2025/327) - Explicitly references the AI Act for AI systems processing health data.
Contributing
Contributions are welcome. Read CONTRIBUTING.md for guidelines.
Please open an issue if you find a broken link or know a resource that belongs here.
To the extent possible under law, Carlos Hernandez has waived all copyright and related rights to this work.
Footnotes
Maintained by Carlos Hernandez (GitHub @carloshvp).