Free management resource

Responsible AI Meeting Checklist

A practical 20-point check for meetings where AI-generated analysis, summaries, recommendations or decisions may influence real work.

AI can help people prepare, analyse and document a meeting, but it should not obscure weak evidence or human accountability. Use this checklist before, during and after an important AI-assisted meeting.

How to use it: Check each statement you can confidently answer “yes” to. Every unchecked item is a reason to pause, clarify or assign an action. Higher-risk meetings may require qualified legal, privacy, HR, compliance, security, safety or other specialist review.

1. Purpose and participation

Make the meeting’s purpose, authority and participants clear.

The meeting purpose, required decision and desired outcome are clearly stated.

The right decision-makers, subject-matter experts and affected perspectives are represented.

Participants know what authority the meeting has—and what requires further approval.

The chair or facilitator is responsible for ensuring that AI supports rather than replaces discussion.

2. AI use and preparation

Be transparent about how AI contributed to the meeting material.

Participants are told where AI was used to prepare analysis, options, slides, summaries or recommendations.

Prompts, source material and important assumptions are available for review where appropriate.

Confidential, personal and restricted information has been handled according to policy.

AI-generated material was reviewed by a knowledgeable person before the meeting.

3. Evidence and discussion

Challenge the evidence instead of accepting confident language.

Important facts, figures, quotations and sources have been independently verified.

Uncertainty, missing information and assumptions are identified rather than presented as facts.

Realistic alternatives, objections, trade-offs and affected stakeholders are discussed.

Participants have time and psychological safety to question the AI-assisted material.

4. Fairness, risk and judgement

Consider who may be affected and what happens if the recommendation is wrong.

Potential bias, exclusion and unequal impact have been considered.

The level of scrutiny matches the legal, financial, employment, safety or reputational consequences.

The final judgement is made by accountable people rather than delegated to the AI system.

The group knows when to defer the decision and obtain specialist advice.

5. Decisions, records and follow-up

Document what was decided and maintain clear human ownership.

The decision record distinguishes evidence, AI-generated suggestions and human judgement.

Actions, owners, deadlines, dependencies and approval points are explicit.

AI-generated minutes or summaries will be checked and corrected before distribution.

There is a plan to monitor results, revisit assumptions and respond if the decision causes unexpected harm.

Your meeting-readiness score: 0/20Pause and strengthen the meeting process.

17–20: Ready to proceedThe meeting process is well structured. Resolve remaining gaps and maintain human review.

12–16: Strengthen safeguardsImportant questions remain. Clarify them before making a consequential decision.

0–11: PauseThe meeting is not ready to rely on AI-assisted material. Rework the fundamentals.

Important: This checklist is a practical meeting aid, not legal, regulatory, privacy, security, HR or safety advice. Review requirements should reflect the decision, participants, information involved and consequences of error.

Frequently asked questions

Should every meeting disclose the use of AI?

Disclosure should be proportionate and meaningful. It is especially important when AI materially shaped analysis, recommendations, records or decisions.

Can AI take meeting minutes automatically?

It can assist, but an accountable person should check accuracy, omissions, confidentiality and agreed actions before distribution.

Does a high score guarantee a responsible decision?

No. The score indicates that common safeguards were considered. It cannot guarantee that the evidence, judgement or final decision is correct.

Turn responsible AI principles into everyday practice

The Practical AI Toolkit for Managers & Professionals includes 500 AI prompts, workflows and decision tools for real work, together with worked examples, review tools, change leadership resources and practical frameworks.

Explore the Practical AI Toolkit

Prefer to start free? Get 50 Executive AI Prompts Every Manager Should Know.