AI readiness resource

AI Readiness Checklist for Leaders

Use these questions to assess business fit, workflow readiness, information risk, team capability, governance, and pilot design before investing further in AI.

Use the checklist online
View all resources

How to use the checklist

Readiness is more than access to a tool.

Work through the questions with the leader who owns the outcome, the people who perform the work, and whoever is responsible for technology, data, security, or compliance.

Ready: true, documented, and understood.

In progress: partly true or dependent on unresolved work.

Not ready or unknown: false, undocumented, or nobody can answer confidently.

Treat the result as a discussion aid, not a compliance certification or guarantee of results.

1. Business purpose

Can we name the problem or opportunity in one clear sentence?
Does a leader own the result and the scope decisions?
What should improve if this works?
Does the use case support a current priority?
What quality, customer, safety, and compliance expectations must remain true?

2. Workflow readiness

Have we selected a specific repeated workflow?
Are its start, finish, owner, inputs, outputs, and handoffs clear?
Where does delay, repetition, rework, or inconsistency occur?
Which steps require judgment and which are repeatable?
Do examples, templates, or quality standards already exist?

3. Information and risk

What information would the AI system receive?
Does it include confidential, personal, regulated, or client-owned data?
Which tools, accounts, and data-handling settings are approved?
Which privacy, security, contractual, or sector requirements apply?
Can early testing use sanitized, synthetic, public, or otherwise approved data?

4. People and adoption

Are the people who perform the work involved in design?
Do they understand AI’s capabilities, limits, and failure patterns?
Who creates, reviews, approves, escalates, and corrects the work?
What workflow-specific practice will participants need?
Has leadership made time for testing, feedback, and documentation?

5. Governance and review

Is one named person accountable for the pilot?
Where must qualified human review occur?
What defines an acceptable output?
How will errors, unexpected behavior, and policy questions be escalated?
Can the pilot be paused or reversed without interrupting essential work?

6. Pilot design and evidence

Is the pilot narrow enough for a defined group and workflow?
Have we recorded a baseline for the current process?
Do success measures cover quality and risk as well as speed?
Are review checkpoints and a go, adjust, or stop decision defined?
What documentation, training, controls, and ownership would scaling require?

Stop and resolve first

Do not pilot around a serious readiness gap.

No leader owns the outcome.
The team cannot identify the information entering the AI system.
Sensitive data would be used without approved tools and handling rules.
High-impact outputs could reach people or systems without qualified review.
No baseline, success measure, escalation path, or rollback exists.
A consequential use has not received appropriate expert review.

Turn unknowns into an action plan

Prioritize the gaps that most affect safety or value.

If the checklist surfaced several unknowns, the AI Readiness Assessment can turn them into a ranked use-case map and practical next steps.

Explore the assessment
Discuss your priorities