Tabular Foundation Models (TFMs) logopeterindia.net
Knowledge Hub

AI Governance Frameworks

AI governance frameworks: regulation, standards, accountable roles and operational controls.

49 topics•5 categories•Updated November 2026

What Is an AI Governance Framework?

15 entries

An AI governance framework is the set of rules, standards, roles and controls that lets an organization scale AI responsibly. A LinkedIn post by Prashant Rathi (31 October 2026) frames it as four connected layers rather than a policy document. Regulation is the legal layer: the EU AI Act, GDPR, CCPA, India's DPDP law, sector-specific rules, cybersecurity laws and emerging AI legislation across regions. Standards structure the work: ISO/IEC 42001, 23894 and 42005 for AI management, risk and impact assessment, and the NIST AI RMF with its Govern, Map, Measure and Manage approach. Roles assign accountability: AI governance boards, model owners, data protection officers, CISOs, auditors, deployers and providers. Controls make it operational: an AI inventory, risk and impact assessments, evals and red teaming, data classification and retention, monitoring of logs, audits and incidents, human oversight, control of agent identities and tool permissions, and least privilege. The post concludes that AI governance is not one framework, law or team but an operating system linking legal obligations, technical safeguards, accountability and continuous oversight. This page follows those layers and links sources for each. Descriptions are paraphrased, are not legal advice, and no compliance claims are made.

01AI governance, simplified (LinkedIn post by Prashant Rathi)The post that describes the four layers of regulation, standards, roles and controls. 02AI Governance and Runtime SecurityGovernance and runtime enforcement platforms, on this site. 03AI Agent Governance PlatformsGoverning autonomous agents, on this site. 04Guardrails for Agentic AIGuardrails for autonomous agents, on this site. 05ComplianceSOC 2, ISO 27001, NIST, GDPR, HIPAA and the EU AI Act, on this site. 06Policy as CodePolicy engines and compliance checks, on this site. 07Data Governance PlatformsData governance tools, on this site. 08AI Explainability TechniquesExplainability methods, on this site. 09LLM Evaluation MethodsEvaluating language models, on this site. 10Hallucination Mitigation MechanismsReducing hallucinations, on this site. 11AI GatewaysRouting, caching and cost governance, on this site. 12AI Agent Security SolutionsSecuring agents, on this site. 13AI Agent ObservabilityObserving agent behavior, on this site. 14Healthcare AI AgentsHealthcare agents and oversight, on this site. 15CyberprivacyPrivacy and cybersecurity, on this site.

Layer 1: Regulation and Policy

14 entries

Laws and policy frameworks that create legal obligations. Not legal advice.

16EU AI Act (European Commission)The European Commission describes the AI Act as the first legal framework on AI, addressing the risks of AI. 17EU AI Act explorer (artificialintelligenceact.eu)Independent site tracking developments in the EU AI Act. 18EU AI Act regulation text (EUR-Lex)The regulation on EUR-Lex; the site returns a non-standard response to automated checks and may need a manual click. 19GDPR (EUR-Lex)The General Data Protection Regulation on EUR-Lex; the site returns a non-standard response to automated checks and may need a manual click. 20European Data Protection BoardThe EU body that issues guidance on applying the GDPR. 21California Consumer Privacy Act (CCPA)The California Attorney General's page on the CCPA, which gives consumers more control over personal information. 22India data protection framework (MeitY)Ministry of Electronics and Information Technology page on India's data protection framework. 23UK: AI regulation, a pro-innovation approachGOV.UK white paper on the UK approach to AI regulation. 24Council of Europe: Framework Convention on AICouncil of Europe page on its AI convention; the site returns 403 to automated checks and may need a manual click. 25OECD AI PrinciplesOECD.AI describes principles for trustworthy AI and a database of national AI policies. 26UNESCO Recommendation on the Ethics of AIUNESCO's global recommendations for ethical AI. 27IAPP Global AI Law and Policy TrackerIAPP tracker of AI legislative and policy developments in a subset of jurisdictions. 28ENISAThe EU agency for cybersecurity, with guidance and tools. 29CISA: AIUS Cybersecurity and Infrastructure Security Agency page on AI.

Layer 2: Standards and Risk Frameworks

8 entries

Voluntary standards and frameworks that structure AI management, risk and impact assessment.

30ISO/IEC 42001ISO page for the AI management system standard; the site returns 403 to automated checks and may need a manual click. 31ISO/IEC 23894ISO page for guidance on AI risk management; the site returns 403 to automated checks and may need a manual click. 32NIST AI Risk Management FrameworkNIST's framework for managing AI risk, organized around Govern, Map, Measure and Manage. 33NIST AI RMF PlaybookNIST Playbook of suggested actions and references for achieving the framework's outcomes. 34NIST AI 600-1: Generative AI ProfileNIST profile that applies the AI RMF to generative AI. 35OWASP GenAI Security ProjectCommunity initiative identifying and tackling the risks of generative AI systems and applications. 36MITRE ATLASKnowledge base of adversary tactics and techniques against AI systems. 37AI Verify FoundationSingapore community for trustworthy AI, with AI governance principles and testing resources.

Layer 3: Roles, Accountability and Documentation

5 entries

Who is accountable, and the artifacts that make accountability auditable.

38IAPPIAPP describes the professional home for privacy, AI governance and digital responsibility. 39Model Cards for Model Reporting (arXiv)Paper proposing model cards that document a model's intended use and performance. 40Datasheets for Datasets (arXiv)Paper proposing documentation for datasets. 41Hugging Face model cardsHugging Face documentation on model cards. 42CycloneDX ML-BOMCycloneDX describes a machine learning bill of materials for model and dataset transparency.

Layer 4: Operational Controls

7 entries

Tools and platforms for testing, monitoring and enforcing controls.

43garakNVIDIA's LLM vulnerability scanner; GitHub lists v0.17.1 (16 October 2026) as the latest release. 44PyRITMicrosoft's Python Risk Identification Tool for generative AI red teaming. 45promptfooOpen-source tool for testing prompts, agents and RAG systems and for AI red teaming; GitHub lists 0.125.1 (25 October 2026). 46Inspect (UK AI Security Institute)Open-source framework for large language model evaluations; GitHub lists 0.3.287 (31 October 2026) for inspect_ai. 47Open Policy AgentPolicy engine for enforcing rules as code, including least-privilege decisions. 48IBM watsonx.governanceIBM describes governing AI across the lifecycle with visibility, control and accountability. 49Credo AICredo AI describes discovering AI, enforcing policies, proving compliance and managing risk.

Frequently Asked Questions

What is an AI governance framework?

The combination of laws, standards, accountable roles and operational controls that lets an organization deploy AI responsibly. A LinkedIn post by Prashant Rathi describes it as four connected layers: regulation, standards, roles and controls.

What are the four layers?

Regulation (for example the EU AI Act, GDPR, CCPA and India's DPDP law), standards (ISO/IEC 42001, 23894 and 42005, and the NIST AI RMF), roles (governance boards, model owners, data protection officers, CISOs, auditors, deployers and providers) and controls (inventory, assessments, testing, monitoring, human oversight and access control).

What is the difference between a regulation and a standard?

Regulations are legally binding, such as the EU AI Act. Standards and frameworks, such as ISO/IEC 42001 and the NIST AI RMF, are voluntary guidance unless a law or contract requires them. See the Compliance page.

What is the NIST AI RMF?

NIST's framework for managing AI risk, built around four functions, Govern, Map, Measure and Manage, with a Playbook of suggested actions and a Generative AI Profile (NIST AI 600-1).

What operational controls does governance need?

The post lists an AI inventory, risk and impact assessments, evals and red teaming, data classification and retention controls, monitoring of logs, audits and incidents, human-in-the-loop oversight, control of agent identities and tool permissions, and least privilege. See the AI Governance and Runtime Security and AI Agent Governance pages.

Why do agents need identity and permission controls?

Agents can call tools and act on systems, so each needs its own identity and narrowly scoped permissions, with least privilege enforced and actions logged. See the AI Agent Governance, Guardrails and Agent Security pages.

Is this legal advice?

No. This page lists sources for orientation. Obligations depend on jurisdiction, sector and use case, so take qualified legal advice.

How current is this page?

Links and release facts were checked on 1 November 2026. The ISO and Council of Europe pages return 403 and the EUR-Lex pages return non-standard responses to automated checks, so those cards may need a manual click. LinkedIn may ask visitors to sign in. Descriptions follow each source's own pages and are not rankings.