Tier A Law · EU
EU AI Act — Regulation (EU) 2024/1689 European Parliament and Council · EUR-Lex The official legal instrument establishing the EU’s risk-based AI framework.
Distinguishes prohibited practices, high-risk systems, transparency and general-purpose AI obligations. Use the articles and annexes to assess applicability; confirm current amendments and implementation guidance. Access: Free legal text
Official instrument confirmed through search retrieval; direct request returned HTTP 202 without readable text.
eur-lex.europa.eu Tier A Guidance · EU
Navigating the AI Act European Commission Official FAQ on scope, risk categories, general-purpose models and implementation.
Clarifies provider and deployer responsibilities, including certain non-EU actors. Organizes implementation questions by topic; the FAQ does not replace the legal text. Access: Free public resource
Official page retrieved and relevant content inspected.
digital-strategy.ec.europa.eu Tier A Framework · US · Global
NIST AI Risk Management Framework 1.0 National Institute of Standards and Technology Voluntary framework for managing AI risks through Govern, Map, Measure and Manage.
Structure ownership, context mapping, evaluation and ongoing risk treatment around four functions. Apply multidisciplinary review across the lifecycle; NIST’s online core notes revision work is underway. Access: Free public resource
Official page retrieved and relevant content inspected.
www.nist.gov Tier A Guidance · US · Global
NIST AI RMF Playbook National Institute of Standards and Technology Suggested actions aligned with RMF outcomes, not a mandatory checklist or prescribed sequence.
Select actions appropriate to each AI use case and operating context. Download PDF, CSV, Excel or JSON for internal workflows; updates will follow the RMF revision. Access: Free public resource
Official page retrieved and relevant content inspected.
airc.nist.gov Tier A Framework · US · Global
NIST AI 600-1 — Generative AI Profile National Institute of Standards and Technology Cross-sector companion extending the AI RMF to generative AI risks and suggested actions.
Address confabulation, information integrity, privacy and other GenAI-specific risk areas. Incorporate trustworthiness into design, use and evaluation rather than relying only on output filtering. Access: Free PDF
Official PDF downloaded successfully; title and publication text inspected.
nvlpubs.nist.gov Tier A Standard · Global
ISO/IEC 42001:2023 — AI management system ISO / IEC Requirements for establishing, implementing, maintaining and continually improving an AI management system.
Organize AI policies, objectives, responsibilities and processes into a management system. Use continual improvement to sustain governance; reading an overview does not establish conformity. Access: Free overview; full standard paid
Official catalogue description retrieved through search; direct access blocked (HTTP 403). Full standard not inspected.
www.iso.org Tier A Standard · Global
ISO/IEC 23894:2023 — AI risk management ISO / IEC Guidance on AI-specific risks and integration of risk management into organizational activities.
Integrate AI risk management with existing business functions and processes. Tailor implementation to the organization and application context. Access: Free abstract; full standard paid
Official abstract retrieved through search; direct access blocked (HTTP 403). Full standard not inspected.
www.iso.org Tier A Framework · Global
OECD AI Principles Organisation for Economic Co-operation and Development Intergovernmental values-based principles and policy recommendations for trustworthy, human-centric AI.
Frame human rights, transparency, robustness and accountability expectations. Use shared vocabulary for policy and supplier discussions; principles are not a compliance certificate. Access: Free public resource
Official page retrieved and relevant content inspected.
oecd.ai Tier A Framework · Global
Recommendation on the Ethics of Artificial Intelligence UNESCO Human-rights-centered ethical values, principles and policy-action areas.
Include fairness, human oversight and environmental sustainability in governance review. Translate principles into policy work on data, gender and wider societal impacts. Access: Free public resource
Official page retrieved and relevant content inspected.
www.unesco.org Tier A Regulator guidance · UK
AI and data protection guidance Information Commissioner’s Office Regulator guidance on data-protection law and good practice when AI processes personal data.
Review fairness, accountability and risks to people from personal-data processing. Under review following the Data (Use and Access) Act; confirm updated guidance before relying on it. Access: Free public resource
Official page retrieved and relevant content inspected.
ico.org.uk Tier A Guidance · UK
Introduction to AI assurance Department for Science, Innovation and Technology Practitioner introduction to assurance concepts, techniques and their role in AI governance.
Build shared terminology for evidence, assurance mechanisms and oversight. Understand how assurance techniques and technical standards support responsible deployment. Access: Free public resource
Official page retrieved and relevant content inspected.
www.gov.uk Tier A Guidance · Singapore · Global
AI Verify Testing Toolkit AI Verify Foundation Testing framework and toolkit combining technical tests with governance-process records for traditional AI.
Test fairness, explainability and robustness with developers and governance teams. Produce customizable reports; this is not a universal GenAI test suite or compliance certification. Access: Free / open-source toolkit
Official page retrieved and relevant content inspected.
aiverifyfoundation.sg Tier A Framework · Singapore
Model AI Governance Framework for Generative AI AI Verify Foundation / IMDA Nine-dimension voluntary framework with initial steps toward a trusted GenAI ecosystem.
Address accountability across the development and deployment lifecycle. Review data quality, safety and trusted development practices alongside other ecosystem dimensions. Access: Free public resource
Official page retrieved and relevant content inspected.
aiverifyfoundation.sg Tier A Framework · India
Responsible AI for All — Part 1: Principles NITI Aayog India-focused approach document establishing responsible-AI principles and societal considerations.
Ground policy discussions in safety, equality, inclusion, privacy and accountability. Use principles as a starting point and pair them with the operationalization companion. Access: Free PDF
Official PDF returned HTTP 200; valid PDF payload confirmed and substantive text retrieved through search.
www.niti.gov.in Tier A Guidance · India
Responsible AI for All — Part 2: Operationalizing Principles NITI Aayog Companion approach document focused on putting responsible-AI principles into practice.
Pair implementation discussions with Part 1’s principles rather than treating them as standalone law. Use the report’s operational approach to inform governance responsibilities and safeguards. Access: Free public resource
Official page retrieved and relevant content inspected.
www.niti.gov.in Tier A Guidance · Canada
Directive on Automated Decision-Making Treasury Board of Canada Secretariat Requirements for covered Canadian federal administrative decision systems, not a general private-sector AI law.
Examine transparency, accountability and procedural-fairness requirements within the directive’s scope. Use linked scope and impact-assessment guidance when deciding which safeguards apply. Access: Free government directive
Substantive official directive retrieved through search; direct response was a Request Rejected page.
www.tbs-sct.canada.ca Tier A Guidance · Canada
Algorithmic Impact Assessment Treasury Board of Canada Secretariat Questionnaire determining impact levels for automated decision systems under the federal directive.
Assess system design, decisions, data and impacts, including mitigation questions. Review scoring and publication instructions; mandatory use is limited to applicable government scope. Access: Free public guidance and open-licensed tool
Official page retrieved and relevant content inspected.
www.canada.ca Tier A Guidance · Australia
Guidance for AI adoption — implementation guidance National AI Centre · Australian Government Six essential practices for organizations using complex or higher-risk AI and needing stronger governance.
Establish accountability, AI registers and supply-chain oversight. Keep records of testing, incidents, monitoring and governance decisions. Access: Free public guidance
Substantive official guidance retrieved through search; direct request timed out.
www.ai.gov.au Tier A Guidance · Global
OWASP Top 10 for LLM Applications OWASP GenAI Security Project Community security guidance on major LLM and GenAI application risks and mitigations.
Threat-model prompt injection, sensitive-information disclosure and unsafe output handling. Review excessive agency; the canonical page displays 2025 while a 2026 release page is indexed. Version-neutral reference pending reconciliation. Access: Free public resource
Official page retrieved and relevant content inspected.
genai.owasp.org Tier A Framework · US · Global
MIT AI Risk Initiative MIT AI Risk Initiative Academic gateway to a living AI risk repository and linked risk, incident and governance datasets.
Broaden risk identification beyond cybersecurity to societal and human impacts. Explore relationships between risks, incidents, mitigations and governance coverage. Access: Free public resource
Official page retrieved and relevant content inspected.
airisk.mit.edu Tier A Guidance · Global
MITRE ATLAS MITRE Knowledge base of adversarial tactics and techniques against AI-enabled systems.
Use technique identifiers to organize threat models and adversarial test plans. Map security testing to realistic attack behaviors; coverage does not prove system safety. Access: Free public resource
Official application returned HTTP 200 and MITRE ATLAS title; interactive catalogue requires browser review.
atlas.mitre.org Tier B Vendor documentation · Global
OpenAI Agents SDK — Guardrails OpenAI SDK patterns for input, output and tool guardrails with tripwires.
Use checks to reject unsafe content or stop an agent run. Review tool and agent guardrail scope; application checks do not replace human approval or broader governance. Access: Free documentation; model usage may incur charges
Official page retrieved and relevant content inspected.
openai.github.io Tier B Vendor documentation · Global
Amazon Bedrock Guardrails Amazon Web Services Configurable safeguards for harmful content, denied topics, sensitive information and grounding.
Apply content filters and probabilistic PII blocking or masking to inputs and responses. Evaluate contextual grounding and safeguard limitations; filtering is not a guarantee of correctness. Access: Free documentation; service usage is charged
Official page retrieved and relevant content inspected.
docs.aws.amazon.com Tier C Advisory · Authored
AEGIS gate thresholds & priority formula AEGIS authored advisory model Workspace gate floors and remediation priorities are consulting assumptions, not external standards.
Gate floors vary by risk tier; priorities combine severity, control weight and risk multiplier. Validate thresholds for your organization before making a release decision. Access: Included authored model
Internal model dated 2026-Q3; no independent external validation.
Authored model · no external link