Back to Frameworks

OECD Due Diligence Guidance for Responsible AI (2026)

International (OECD)
v2026 (approved 26 January 2026; revised version May 2026)
7 domains
24 controls

The OECD's 2026 guidance on due diligence for responsible AI: how suppliers of AI inputs, developers and deployers, and users of AI systems implement the MNE Guidelines' six-step due diligence and the OECD AI Principles, from policies and management systems through risk scoping, assessment, involvement and prioritisation, mitigation (data sourcing, transparency, security, responsible deployment, workplace AI), value chain leverage and investor engagement, to tracking, disclosure and remediation, with meaningful stakeholder engagement throughout.

Verified

OECD Due Diligence Guidance for Responsible AI (2026) is a compliance framework from International (OECD) with 7 domains and 24 controls. The largest domains are Step 2: Identify and assess actual and potential adverse impacts – OECD Due Diligence Guidance for Responsible AI (2026) (8 controls), Step 3: Cease, prevent and mitigate adverse impacts – OECD Due Diligence Guidance for Responsible AI (2026) (8 controls), Step 1: Embed responsible business conduct into policies and management systems – OECD Due Diligence Guidance for Responsible AI (2026) (4 controls). Every control below carries what it requires and what an assessor expects to see.

Maintained by Gerard Blokdyk

Framework summaries on this platform are AI-assisted interpretations for educational and compliance planning purposes. They do not reproduce or replace the official standards. Refer to the authoritative source for the definitive text. Framework names and trademarks belong to their respective organisations.

Framework Domains (7)

Meaningful stakeholder engagement (all steps) – OECD Due Diligence Guidance for Responsible AI (2026)

1 controls
Controls in the Meaningful stakeholder engagement (all steps) – OECD Due Diligence Guidance for Responsible AI (2026) domain of OECD Due Diligence Guidance for Responsible AI (2026) — 1 controls
CodeTitle
oecd-due-diligence-guidance-for-responsible-ai-2026::SESE Meaningful stakeholder engagement throughout due diligence

Step 1: Embed responsible business conduct into policies and management systems – OECD Due Diligence Guidance for Responsible AI (2026)

4 controls
Controls in the Step 1: Embed responsible business conduct into policies and management systems – OECD Due Diligence Guidance for Responsible AI (2026) domain of OECD Due Diligence Guidance for Responsible AI (2026) — 4 controls
CodeTitle
oecd-due-diligence-guidance-for-responsible-ai-2026::1.11.1 Responsible business conduct policies for AI
oecd-due-diligence-guidance-for-responsible-ai-2026::1.2.1-71.2.1-7 Management systems: oversight, responsibilities, records, communication and incentives
oecd-due-diligence-guidance-for-responsible-ai-2026::1.2.8-161.2.8-16 Management systems: process integration, incidents, grievances, decommissioning and contingency
oecd-due-diligence-guidance-for-responsible-ai-2026::1.31.3 Expectations on business relationships

Step 2: Identify and assess actual and potential adverse impacts – OECD Due Diligence Guidance for Responsible AI (2026)

8 controls
Controls in the Step 2: Identify and assess actual and potential adverse impacts – OECD Due Diligence Guidance for Responsible AI (2026) domain of OECD Due Diligence Guidance for Responsible AI (2026) — 8 controls
CodeTitle
oecd-due-diligence-guidance-for-responsible-ai-2026::2.1.G12.1.G1 Initial scoping of risks: suppliers of AI inputs
oecd-due-diligence-guidance-for-responsible-ai-2026::2.1.G22.1.G2 Initial scoping of risks: enterprises in the AI system lifecycle
oecd-due-diligence-guidance-for-responsible-ai-2026::2.1.G32.1.G3 Initial scoping of risks: users of AI systems
oecd-due-diligence-guidance-for-responsible-ai-2026::2.2.G1G32.2.G1G3 In-depth assessment by suppliers and users of high-risk AI developers
oecd-due-diligence-guidance-for-responsible-ai-2026::2.2.G2-own2.2.G2-own In-depth assessment of significant risks in own operations: enterprises in the lifecycle
oecd-due-diligence-guidance-for-responsible-ai-2026::2.2.G2-rel2.2.G2-rel In-depth assessment of risks through business relationships using the enterprise's AI
oecd-due-diligence-guidance-for-responsible-ai-2026::2.32.3 Assess involvement: cause, contribute or directly linked
oecd-due-diligence-guidance-for-responsible-ai-2026::2.42.4 Prioritise the most significant risks

Step 3: Cease, prevent and mitigate adverse impacts – OECD Due Diligence Guidance for Responsible AI (2026)

8 controls
Controls in the Step 3: Cease, prevent and mitigate adverse impacts – OECD Due Diligence Guidance for Responsible AI (2026) domain of OECD Due Diligence Guidance for Responsible AI (2026) — 8 controls
CodeTitle
oecd-due-diligence-guidance-for-responsible-ai-2026::3.1.G2-data3.1.G2-data Mitigation by developers: responsible data sourcing and training
oecd-due-diligence-guidance-for-responsible-ai-2026::3.1.G2-deploy3.1.G2-deploy Mitigation by developers: responsible deployment
oecd-due-diligence-guidance-for-responsible-ai-2026::3.1.G2-sec3.1.G2-sec Mitigation by developers: security, safety and robustness
oecd-due-diligence-guidance-for-responsible-ai-2026::3.1.G2-transp3.1.G2-transp Mitigation by developers: transparency, explainability and traceability
oecd-due-diligence-guidance-for-responsible-ai-2026::3.1.G33.1.G3 Mitigation by users of AI systems
oecd-due-diligence-guidance-for-responsible-ai-2026::3.1.all3.1.all Cease and prevent impacts the enterprise causes or contributes to
oecd-due-diligence-guidance-for-responsible-ai-2026::3.23.2 Address risks directly linked through business relationships
oecd-due-diligence-guidance-for-responsible-ai-2026::B2.16B2.16 Due diligence by investors and financial institutions in AI

Step 4: Track implementation and results – OECD Due Diligence Guidance for Responsible AI (2026)

1 controls
Controls in the Step 4: Track implementation and results – OECD Due Diligence Guidance for Responsible AI (2026) domain of OECD Due Diligence Guidance for Responsible AI (2026) — 1 controls
CodeTitle
oecd-due-diligence-guidance-for-responsible-ai-2026::44 Track implementation and results

Step 5: Communicate actions to address impacts – OECD Due Diligence Guidance for Responsible AI (2026)

1 controls
Controls in the Step 5: Communicate actions to address impacts – OECD Due Diligence Guidance for Responsible AI (2026) domain of OECD Due Diligence Guidance for Responsible AI (2026) — 1 controls
CodeTitle
oecd-due-diligence-guidance-for-responsible-ai-2026::55 Communicate actions to address impacts

Step 6: Provide for or co-operate in remediation – OECD Due Diligence Guidance for Responsible AI (2026)

1 controls
Controls in the Step 6: Provide for or co-operate in remediation – OECD Due Diligence Guidance for Responsible AI (2026) domain of OECD Due Diligence Guidance for Responsible AI (2026) — 1 controls
CodeTitle
oecd-due-diligence-guidance-for-responsible-ai-2026::66 Provide for or co-operate in remediation

What is OECD Due Diligence Guidance for Responsible AI (2026) and who does it apply to?

OECD Due Diligence Guidance for Responsible AI (2026) is a compliance framework from International (OECD) with 7 domains and 24 controls. The OECD's 2026 guidance on due diligence for responsible AI: how suppliers of AI inputs, developers and deployers, and users of AI systems implement the MNE Guidelines' six-step due diligence and the OECD AI Principles, from policies and management systems through risk scoping, assessment, involvement and prioritisation, mitigation (data sourcing, transparency, security, responsible deployment, workplace AI), value chain leverage and investor engagement, to tracking, disclosure and remediation, with meaningful stakeholder engagement throughout. It is used by organisations to establish and maintain compliance with industry standards and regulatory requirements.

What does OECD Due Diligence Guidance for Responsible AI (2026) actually require?

OECD Due Diligence Guidance for Responsible AI (2026) has 24 controls organised across 7 domains. The largest domains are Step 2: Identify and assess actual and potential adverse impacts – OECD Due Diligence Guidance for Responsible AI (2026) (8 controls), Step 3: Cease, prevent and mitigate adverse impacts – OECD Due Diligence Guidance for Responsible AI (2026) (8 controls), Step 1: Embed responsible business conduct into policies and management systems – OECD Due Diligence Guidance for Responsible AI (2026) (4 controls). Each control defines specific requirements that organisations must implement to achieve compliance.

If I already comply with another framework, how much of OECD Due Diligence Guidance for Responsible AI (2026) do I already cover?

OECD Due Diligence Guidance for Responsible AI (2026) does not currently have cross-framework mappings in our system. Check back as we continuously expand our mapping database.

How do I implement OECD Due Diligence Guidance for Responsible AI (2026)?

Start your OECD Due Diligence Guidance for Responsible AI (2026) compliance journey by running a self-assessment on our platform to identify your current compliance posture. Our AI advisory can answer specific questions about OECD Due Diligence Guidance for Responsible AI (2026) requirements, and cross-framework mapping helps you leverage existing controls from other frameworks you may already comply with. Create a free account to access all 24 controls and track your progress.

Start Your Compliance Journey

Create a free account to run self-assessments, get AI advisory, and track your compliance progress across 705 frameworks.

Get Started Free →

Free forever — no credit card required