STEP 01 OF 06
4Ps Self-Assessment
Twenty questions, score each from 1 (not yet) to 5 (fully in place). The radar updates live. When you finish, the result panel on the right tells you your biggest gap and your next step.
Progress: 0/20 (0%)
PrimedOur executive team can describe our top three AI use cases in plain language.
1 not yet · 5 fully in place
PrimedWorkforce literacy in AI is trained role-by-role, not as a generic course.
1 not yet · 5 fully in place
PrimedData each priority use case depends on is documented and fit for use.
1 not yet · 5 fully in place
PrimedOur AI platform stack is documented, with cost and access controls in place.
1 not yet · 5 fully in place
PrimedWe have a written list of priority AI use cases, each with a stated purpose, owner, success measure and risk class.
1 not yet · 5 fully in place
PrincipledWe have a short, public set of AI principles that the team can recite.
1 not yet · 5 fully in place
PrincipledFor each material AI decision, the RACI is published.
1 not yet · 5 fully in place
PrincipledWe run an ethics review with a chair, a quorum, and a cadence.
1 not yet · 5 fully in place
PrincipledWe have clear rules for when AI use must be disclosed externally and internally.
1 not yet · 5 fully in place
PrincipledHuman-in-the-loop rules are documented per use case.
1 not yet · 5 fully in place
PractisedAI use cases follow a defined lifecycle with entry/exit criteria.
1 not yet · 5 fully in place
PractisedEach production model has a defined lifecycle including red teaming pre-deploy.
1 not yet · 5 fully in place
PractisedWe run AI work to a deliberate operating cadence (weekly/monthly/quarterly).
1 not yet · 5 fully in place
PractisedThe team has the People/Purpose/Principles/Practices/Productivity/Performance (6Ps) in place.
1 not yet · 5 fully in place
PractisedVendor and partner relationships are managed against a written scorecard.
1 not yet · 5 fully in place
ProtectedOur AI risk inventory is current with named owners and mitigations, including vendor compute supply and concentration risk.
1 not yet · 5 fully in place
ProtectedWe have a written mapping to EU AI Act / NIST AI RMF / ISO 42001.
1 not yet · 5 fully in place
ProtectedEvery material production model has a current model card.
1 not yet · 5 fully in place
ProtectedWe can reconstruct an AI-mediated decision from our audit trail within an hour.
1 not yet · 5 fully in place
ProtectedOur incident response playbook has been rehearsed in the last twelve months, including a vendor capacity-rationing or model-withdrawal scenario (a provider or government restricting a model, as happened in June 2026) with a fallback we control.
1 not yet · 5 fully in place