AI project readiness check

An AI-enabled project may have a sponsor and a technical route while still lacking an accountable owner, suitable data, acceptance evidence, human oversight or a workable fallback. This check makes those gaps visible before the next decision.

About 6 minutesNo sign-upNo answers saved or sentNo readiness score

Readiness evidence

0 of 12 checked

01Business objective and case for AIThe problem, affected users, expected outcome and reason for using an AI-enabled approach are clear enough to test.

Clear and evidencedNamed owner, current answer and proportionate evidencePartly clearSome useful work exists, but gaps or conflicts remainNot yet clearUnknown, assumed, outdated or without an ownerNot applicableNot relevant to this context, with a reason recorded

02Accountable ownershipA named business or service owner is accountable for the outcome, affected people and accepted residual risk.

03Affected people and potential impactThe project has identified users and other affected people, including who may experience an error or decision differently.

04Data suitability, permission and ownershipData for development, evaluation and live use has an owner, known provenance, suitable coverage and a permitted route for use.

05Legal, compliance and privacy routeThe relevant legal, compliance, privacy and sector specialists are involved early enough to influence the design and decisions.

06Security and accessThe system boundary, information flows, access rights, integrations and credible misuse have been identified for security review.

07Model, build, buy and supplier dependenciesThe delivery route and dependencies are visible, including models, suppliers, hosting, terms, change controls, portability and exit needs.

08Evaluation and acceptance evidenceMeasures, tolerances, representative test cases, known limitations and approval evidence are defined before the result is known.

09Human oversight, override and escalationHuman review has defined authority, competence, time, information, escalation and an effective route to override or stop the capability.

10Operational ownership, skills and changeThe future process, roles, workload, training, support and operating owner are defined beyond the technical implementation.

11Monitoring, incidents and reviewNamed owners, indicators, thresholds and review points are defined for live performance, errors, impact, incidents and material change.

12Fallback, exit and decommissioningThere is a credible, owned and funded route to restrict, roll back, replace or retire the capability if it is unsuitable or unavailable.

Complete all 12 checks to see the evidence gaps.The result will list every area marked Not yet clear or Partly clear. It will not calculate a readiness score.

How to use the result

Turn gaps into owned work and evidence

Take the unanswered and partly answered areas into the next sponsor, governance or delivery discussion. Give each material question an owner, a next action, an evidence location and a review date.

The check does not determine legal applicability, security, data suitability or model fitness. Project managers should make those decisions visible and timely, while the relevant specialists provide the judgement and approval required by local governance.

Connected working resources

Readiness questions lead into risk and governance work.

The AI project delivery hub connects this check to the risk guide, governance checklist and editable Excel workbook.

AI project delivery toolkit

Practical project tools shaped by more than 25 years of project work. AI is used to support production. Content is co-produced, edited and quality-checked by the site owner.