An AI policy or charter sets out the organisation’s expectations, commitments, and governance arrangements for the responsible use of AI. This practice builds on Define Responsible AI Principles by turning those values into clear, actionable guidance for teams.
The policy typically outlines how those principles are applied in practice, including roles, decision-making processes, escalation pathways, and links to relevant governance mechanisms. A well-crafted AI policy provides a foundation for consistent decision-making and helps align AI use with the organisation’s goals as adoption grows.
Why it matters
An AI policy provides a single reference point for teams working with or impacted by AI. It builds a shared understanding of the guardrails for responsible use, reduces the risk of siloed or informal practices, and signals the organisation’s commitment to accountability, internally and externally. It is a key artefact to start demonstrating readiness to regulators, customers, and partners.
But it is not a standalone solution, it is one part of a broader AI governance approach. To be effective, it needs to be supported by other practices like risk assessment, human oversight, system inventory, capability building, and procurement governance. Together, these elements help ensure a policy is not just words on paper but part of a living, actionable framework for responsible AI use.
Implementation tips
- Leverage existing policies as a starting point (e.g. privacy or risk policies).
- Involve a cross-functional group to draft the policy to reflect a range of perspectives and ensure broad buy-in and relevance.
- Make the policy actionable by linking principles to expected behaviours, controls, or decisions..
- Use plain language and visuals where possible to increase accessibility and engagement.
- Consider shorter review cycles to cater for changing regulations, technologies, or uses of AI.
- If your organisation is using generative AI tools, consider whether to address usage expectations within the AI policy or in a separate Acceptable Use Policy. Keep the distinction clear and consistent.
- Where environmental factors are material to your organisation, consider referencing environmental considerations in your AI policy and linking to relevant environmental assessment processes.
Support materials
Responsible AI Institute – AI Policy Template
A practical starting point for organisations drafting a foundational AI policy.
AI Institute. (2024, January). AI Policy Template. AI Institute.
A downloadable template to guide policy structure and content.
Mott MacDonald – Responsible AI Policy
A real-world corporate example from Mott MacDonald, developed to support responsible AI use in infrastructure and engineering contexts.
Doing AI Governance – Creating Your AI Governance Policy
A step-by-step guide to writing and refining your AI governance policy.
European Institute for Sustainable IT (ISIT) Sustainable AI Charter
One-pager charter focused on sustainability aspects of Artificial Intelligence.
