AI readiness assessment

Use this checklist in a 60–90 minute leadership session. A future interactive version may ship on this URL; for now, copy sections into your notes app or print.

Strategy and outcomes

• Named executive sponsor and decision rights
• 3–5 business outcomes tied to metrics (not “use AI”)
• Explicit non-goals (what you will not automate yet)

Data and systems

• Inventory of systems that hold customer, financial, or IP data
• Data classification (public / internal / restricted)
• Known gaps in APIs, exports, or lineage documentation
• Retention and deletion policies understood for model training use no

Risk and compliance

• Legal/compliance review for regulated data in prompts
• Vendor DPAs and subprocessors reviewed for chosen models
• Incident response path if a model leaks sensitive context
• Human review gates for customer-facing outputs

People and process

• Roles defined: product owner, engineering, security, content
• Training plan for safe prompting and secret handling
• Change management for affected teams

Technology

• Build vs buy documented with exit criteria
• Staging environment for agent experiments (no prod keys)
• Evaluation set for quality before launch—not vibe checks only

Next step

Turn results into AI roadmap template rows with owners and dates. For delivery help, see services.