How AI ready is your organisation?

How actively does your senior management support AI initiatives?

Leadership commitment is the strongest predictor of AI programme success. Without budget, sponsorship and strategic priority from the top, AI initiatives stay grassroots experiments that fade when the enthusiasm does. This question measures where your senior management stands, from no engagement to board-level strategic priority.

What is the level of AI understanding across your organisation?

Whether AI tools get adopted or ignored depends on the wider workforce understanding what AI can do; a small AI team cannot transform an organisation where everyone else is unsure. This question measures how far AI understanding and training have spread. The percentage thresholds in the options keep the answer measurable rather than anecdotal.

What AI talent does your organisation have?

AI Literacy (question 2) is about everyone; this question is about specialists, the people whose job is to build, deploy and maintain AI systems. We measure the depth of dedicated AI talent, from no AI roles at all through to a full in-house centre of excellence.

How is your organisation adapting its workforce structure for AI?

Hiring AI Talent (question 3) is not enough; the existing workforce has to adapt too. AI takes over coordination-heavy work and amplifies human judgement, so roles and team structures will change whether you plan for it or not. This question measures whether your organisation is redesigning roles deliberately, or waiting to be disrupted.

How does your organisation approach AI experimentation?

Adopting AI requires structured experimentation. Unguided pilots with no tracking burn resources and teach nothing, while a pipeline from pilot to production turns experiments into value. We measure how mature your experimentation process is, not simply whether experiments happen.

What AI governance structures exist in your organisation?

Governance is the structural foundation for responsible AI: policies, clear accountability, and review processes. Without them, good intentions stay aspirational. This question measures how far governance has progressed, from nothing in place through to a framework aligned with recognised standards such as the NIST AI Risk Management Framework.

How does your organisation manage AI-related risks?

AI Governance (question 6) sets the rules; this measures whether you can enforce them: the operational ability to identify, assess and mitigate specific AI risks. We look for how far your controls go, from basic awareness through continuous monitoring to independent testing of AI systems before and after rollout.

How does your organisation manage GenAI and AI agent risks (hallucinations, IP exposure, data leakage, autonomous actions)?

Generative AI and agents bring risks that AI Risk Control (question 7) was never designed for: made-up answers, confidential information leaking through prompts, and agents taking actions nobody reviewed. This question measures the controls you have for these specifically, from approved tools and limits on what agents may access to monitoring of what they do.

How well-defined are your AI use cases and value propositions?

AI value starts with clearly defined use cases. This question measures how far yours have progressed, from none identified to a documented portfolio. From the middle options onward it also separates doing existing work faster from restructuring how the work is done, the distinction where the larger gains sit.

How effectively is your organisation capturing productivity gains from GenAI tools?

Defining use cases is not the same as capturing value, and Business Use Case (question 9) only asked what you have on paper. Many organisations roll out tools and never measure the impact. This question measures whether the gains are real and tracked, and at the top end, whether work has been reorganised around AI rather than just done faster.

What is the quality and accessibility of your organisation’s data for AI?

Poor-quality data locked in silos is the most common reason AI projects never get past the pilot. This question measures whether your organisation’s data is AI-ready: consistent, accessible, and covering both structured records and the documents and text that today’s AI can work with.

How mature is your master/reference data management?

Reference data, such as customer lists, product catalogues and reporting structures, is the backbone of consistent AI output. If every system defines "customer" differently, AI trained on that data will be inconsistent too. This measures a capability beyond general data quality (Data Quality, question 11): whether your organisation keeps one trusted version of its core data.

What AI/ML infrastructure and data architecture capabilities does your organisation have?

Running AI in production takes platforms to build and operate on, and data architecture to feed them. This question measures the full technical stack, from nothing in place through to enterprise-grade operations with integrated data pipelines. On-premise, cloud or hybrid, all are equally valid at every level.

How mature are your organisation’s specification standards for AI work delegation?

The organisational counterpart of Specification Thinking in pAIRI, our personal assessment: individual skill has limited impact without shared standards that make it repeatable. We look for instruction and prompt libraries, delegation templates, quality checkpoints and standard operating procedures. That is the difference between a few skilled individuals and an organisation that delegates work to AI reliably.

How advanced is your organisation’s GenAI tool and AI agent deployment capability?

The capstone of the Infrastructure & Standards pillar. Deploying generative AI and agents at scale takes more than the platforms covered by AI/ML/Data Infrastructure (question 13): it takes enterprise tool management, scoped permissions, usage monitoring, and governance for workflows that run on their own. This question measures whether your organisation can deploy and manage them at scale.