AI readiness checklist
Twenty-five things we check in the first week of every assessment. Tick what is true today; your answers stay in this browser. Print it, share it with your leadership team, or send us the result.
0 / 25
Print or save as PDF Reset
Strategy & ownership 0 of 5
A senior leader owns AI decisions, budget and risk. We have a written list of candidate use cases ranked by value and feasibility. Each shortlisted use case has a named business owner. We can say in numbers what success looks like for the first use case. Leadership has agreed how much to invest in the next twelve months.
Data 0 of 5
We know where the data for our top use case lives and who controls it. An engineer could get access to that data within days, not months. We have a sample of real cases that could become a test set. Personal data is classified and we know the rules for using it. Known data-quality problems are written down, even if not yet fixed.
People & skills 0 of 5
At least one engineer can integrate an API and deploy a service. Business users of the first use case can give feedback every week. Someone in the team can judge whether an AI output is correct in that domain. Staff have had basic training on what generative AI can and cannot do. We know what we will build in-house and what we will buy or outsource.
Technology & security 0 of 5
Our IT policy says which model providers we may use. We know our data-residency constraints: in-country, on-premises or VPC. Single sign-on and role-based access exist for internal tools. There is a path to production: environments, CI and monitoring. There is a budget line for model and infrastructure usage, not only for the build.
Governance & risk 0 of 5
We know which regulations apply: DPDP Act, EU AI Act, sector rules. There is an AI use policy that employees have actually read. High-risk uses (credit, hiring, health) are identified and reviewed separately. Outputs can be tested before release, even if the process is manual today. Someone reviews incidents and complaints that involve AI outputs.
