How to Write a Code of Conduct for AI Use at Work

An AI usage policy tells your team what they can and cannot do with AI tools. A code of conduct for AI goes one layer deeper: it articulates the values and principles your organisation applies when using AI, not just the rules. It answers 'why' rather than just 'what'. For leadership communications, ethics frameworks, and external stakeholder commitments, the code-of-conduct format often resonates better than a rulebook.

How it differs from an AI usage policy

An AI usage policy is operational: it specifies approved tools, data handling rules, and reporting mechanisms. A code of conduct for AI is principled: it sets out your organisation's stance on fairness, transparency, human oversight, and the responsible use of AI. In practice, most organisations benefit from both, a policy for staff to follow day to day, and a code of conduct that gives those rules a coherent rationale that people can remember.

Five principles worth including

Human oversight: AI outputs require human review before use in important decisions or external communications. Transparency: be open about AI involvement in your work where it is material. Fairness: monitor AI use for discriminatory patterns, especially in hiring, customer decisions, or content creation. Privacy: only process data through AI tools with appropriate contractual and technical safeguards. Accountability: one named person is responsible for how AI is used across the organisation.

Tone and audience

A code of conduct for AI is often shared more widely than a usage policy, with customers, partners, investors, or prospective hires. The tone should be aspirational and values-led, not legalistic. Write it in the first person plural ('we believe', 'we commit to') and keep it to one page or under. It should read as a genuine commitment, not a liability disclaimer drafted by someone who has never used an AI tool.

Making it real rather than decorative

The risk with any principles document is that it becomes a poster nobody reads. Tie your code of conduct to the actual rules in your AI usage policy by cross-linking the two documents directly. Include it in onboarding materials. Reference it when making AI-related decisions as a team. A code of conduct that shapes actual behaviour is worth writing; one that sits in a drawer is not. ModelCharter's policy generator produces both a detailed usage policy and a values-level summary that functions as a code of conduct.