The AIRI Framework Organisational AI Readiness Index Assessment oAIRI – Organisational AI Readiness IndexStep 1Step 2Step 3Step 4Step 5Step 6Step 7Step 8The Organisational AI Readiness Index (oAIRI) measures where your organisation is today across five pillars — Leadership & Culture · Ethics & Governance · Business Value · Data Foundation · Infrastructure & Standards — so leaders can see exactly where to invest next. Who should answer: Leaders or senior practitioners who can speak to how the organisation operates as a whole. If you can’t assess a dimension honestly, pick the level that matches what you actually see — not what you wish were true. Multiple respondents recommended. Ask 3–5 leaders to complete oAIRI independently and compare. The spread of answers is itself a finding. Reassess every 6 months.PreviousNextPersonalisation step before the assessment begins What happens: after you submit oAIRI, we read your public homepage and about page — the same content visible to any visitor — and produce a short industry profile and potential AI Use Cases your organisation can embark on. Privacy: no login data, no internal documents, no employee or customer data — only your public website. Full NameEmailCompany NameYour organisation’s website I have read and agree to the Terms and Conditions and Privacy PolicyPreviousNextPillar 1 of 5 — Leadership & Culture What this pillar measures: Leadership commitment, workforce literacy, AI talent, workforce adaptation, and experimentation culture — the people and culture foundations that determine whether technical AI investments succeed or fail. Why it matters: 87% of AI projects never make it to production, and the reasons rarely involve technology — they involve leadership commitment, literacy, and culture. This is the execution multiplier. D1. Management Support Strong AI results start at the top. This is about how much your senior leaders back AI — with attention, budget, and priority — rather than leaving it to scattered experiments.How actively does your senior management support AI initiatives? No discussion or budget allocation for AI initiatives Leadership mentions AI occasionally but no dedicated budget exists Active interest with pilot budget approved (>S$50,000 allocated) Strong support — a dedicated AI budget, people and resources, with AI goals built into how performance is measured AI is a board-level strategic priority with multi-year investment roadmapD2. AI Literacy AI tools only get used if people understand them. This is about how widely your employees grasp what AI can do for their work, not just a handful of specialists. What is the level of AI understanding across your organisation? Most employees have no understanding of AI capabilities Basic awareness exists, some employees have tried GenAI chatbots personally >30% of employees have completed AI training or use AI tools regularly Organisation-wide AI training with >60% completion and measured application AI literacy is part of the culture — most employees use AI in their daily work, with ongoing training as tools changeD3. AI Talent This is about whether your organisation has people with the skills to build, run, and look after AI systems — from none, through to a dedicated team.What AI talent does your organisation have? No AI-specific talent or roles Some IT staff with basic AI awareness but no specialists Small AI team (1–5 people) or dedicated AI champions in departments An established AI team (6+ people), including specialists who build and run AI systems A full in-house AI centre with research, building, and product skillsD4. Workforce Adaptation As AI takes on routine work, roles need to change. This is about whether your organisation is actively redesigning roles and teams for that shift, rather than waiting to be caught out.How is your organisation adapting its workforce structure for AI? No assessment of how AI affects roles or team structures Leadership acknowledges AI will change roles but no formal review has been conducted Organisation has audited key roles for AI impact and identified coordination-heavy roles at risk of transformation Active role redesign underway with transition plans, reskilling programmes, and new role definitions that emphasise human judgement Continuous workforce architecture programme that regularly realigns roles, teams, and value chains around evolving AI capabilitiesD5. Experimentation Culture This is about how your organisation tries AI out — whether pilots are run in a planned, tracked way that leads to real adoption, or stay as one-off experiments that go nowhere.How does your organisation approach AI experimentation? No AI experimentation or pilots Employees try AI on their own, but there’s no list of trials, no success measures, and no tracking Structured pilot process with 2–5 pilots completed and learnings documented A clear process for trying AI, with 6 or more trials completed and a way to scale up what works Trying and improving AI is part of everyday operations, with successful trials quickly put to real usePreviousNextPillar 2 of 5 — Ethics & Governance What this pillar measures: Your AI governance structures, how you manage AI-related risk, and how you specifically handle GenAI and AI agent risks (hallucinations, IP exposure, data leakage, autonomous actions). Why it matters: This pillar is the ethics gate for any AI initiative. As AI moves from tools to agents that act autonomously, governance stops being a compliance exercise and starts being a safety requirement.D6. AI Governance Responsible AI needs clear rules. This is about whether your organisation has the policies, ownership, and review processes that keep AI use safe and accountable.What rules and oversight does your organisation have for using AI responsibly? No AI rules or policies in place Ad-hoc decisions without formal structure, considering creating guidelines Basic AI guidelines exist, with a rule that a person reviews important decisions Established AI oversight with clear policies, accountability, and regular reviews A complete, well-recognised set of AI rules in line with leading national or international standards, putting us among the leaderseD7. AI Risk Control This is about how well your organisation spots and manages AI risks in practice — turning policies into real checks and controls that catch problems before they cause harm.How does your organisation manage AI-related risks? No consideration of AI risks Aware of risks but no formal management or documented controls Basic risk identification with some controls and incident response process Systematic risk assessment with documented mitigations and regular audits A complete risk approach with ongoing monitoring and thorough, regular testing of AI before and after it’s usedD8. GenAI & Agent Risk Management Generative AI and AI agents bring their own risks — made-up answers, leaked information, and actions taken without anyone checking. This is about whether your organisation has controls aimed specifically at these.How does your organisation manage the risks from AI chatbots and AI agents — like made-up answers, leaked information, and actions taken without anyone checking? No awareness of GenAI or agent-specific risks, employees use public AI tools and agents freely Aware of GenAI and agent risks but no specific controls; informal guidance only A basic policy for using AI chatbots and agents — an approved-tools list, rules on what AI agents may do, and staff training Formal policies with controls that stop sensitive data leaving, approved company tools, limits on what AI agents can reach (which folders and services), and monitoring of what they do A complete approach — automatic controls, full records of what AI agents do, detection of unapproved AI use, and rules for tasks AI runs on its ownPreviousNextPillar 3 of 5 — Business Value What this pillar measures: How well the organisation identifies high-value AI use cases and how effectively it captures productivity gains from GenAI tools already in use. Why it matters: A strong P3 means the organisation doesn’t just use AI — it redesigns work around AI. The bar has moved from “we saved 2 hours/week” to “we restructured the workflow so humans focus on judgement and AI handles production.D9. Business Use Case AI value starts with knowing where to apply it. This is about whether your organisation has clearly defined, worthwhile uses for AI — and can tell apart simply doing work faster from rethinking how the work is done.How well-defined are your AI use cases and value propositions? No clear AI use cases identified AI uses discussed in meetings but none written down with a business reason or a check on whether the data exists 2–3 use cases documented with basic value estimates and data requirements identified A set of AI uses covering both doing existing work faster and rethinking how work is done, each with clear expected benefits and leadership backing A full set of AI uses with proven benefits; the organisation measures value from genuinely changing how work is done, not just time savedD10. GenAI Value Realisation Buying AI tools is easy; getting value from them is harder. This is about whether your organisation can show real, measured benefit from the AI it has rolled out.How effectively is your organisation capturing productivity gains from GenAI tools? No AI chatbots in use across the organisation Some employees trying everyday AI tools on their own GenAI tools adopted in some departments with anecdotal productivity reports AI chatbots used across the organisation with measured time savings, and we’re starting to track the value of reorganising work so AI does the routine part and people focus on judgement AI built into core ways of working, with recorded benefits from both time saved and reorganised work; teams are redesigned around people and AI working togetherPreviousNextPillar 4 of 5 — Data Foundation What this pillar measures: The quality and accessibility of the organisation’s data for AI, and the maturity of master/reference data management. Why it matters: An AI project can have strong leadership backing, clear value, and clean ethics — and still fail because the data isn’t there or is too fragmented to use. In 2026, “data” means both structured data and documents, emails, and knowledge bases that feed RAG and GenAI systems.D11. Data Quality AI projects most often stall on poor data. This is about whether your organisation’s data is clean, joined-up, and easy enough to reach for AI to use.What is the quality and accessibility of your organisation’s data for AI? Poor data quality, stuck in separate systems, with nothing ready for AI to use Data exists in multiple systems but has known issues (duplicates, gaps, inconsistent formats) and requires manual effort to access for AI use Key business data is consistent and tidy, AND documents are organised so AI can search and use them Good quality with centralised data management for both structured and unstructured data Excellent quality, with data brought together in one place and organised so AI can use it easilyD12. Reference Data Shared reference data — like customer lists, product catalogues, and org charts — keeps AI results consistent. This is about whether your organisation keeps this core data standardised and reliable.How mature is your master/reference data management? No reference data management, inconsistent definitions everywhere Inconsistent data definitions across systems, ad-hoc fixes Some standardisation efforts underway with documentation started Well-managed core data with clear ownership and one trusted version for key things like customers and products Core data managed consistently across the whole organisation, with quality checked continuouslyPreviousNextPillar 5 of 5 — Infrastructure & Standards What this pillar measures: Your AI/ML infrastructure and data architecture, the maturity of specification standards (how people delegate work to AI), and your GenAI/agent deployment capability. Why it matters: Infrastructure alone is no longer enough — in the agent era, specification standards matter just as much. An organisation where everyone writes their own ad-hoc prompts will never scale AI like one with shared prompt libraries, SOPs, and quality gates.D13. AI/ML/Data Infrastructure This is about the technology your organisation has in place to run AI — the computing power and the data pipelines that feed it — from none, through to a full, enterprise-ready setup.What technology setup does your organisation have to run AI and manage its data? No setup for running AI, and no access to AI tools Basic tools like spreadsheets and dashboards, but nothing to run AI; getting at the data takes a lot of manual work Company access to ready-made AI services, or basic AI tools used manually; data systems can feed AI through standard connections A proper system to run AI, with data flowing into it automatically from central stores A complete, professional setup that handles building, improving, and running AI from start to finish, fed by one reliable, up-to-the-minute data system with quality controlsD14. Specification Standards & Practices Individual skill at directing AI only scales if the organisation makes it repeatable. This is about whether your organisation has shared templates, standards, and quality checks for how people set tasks for AI.How mature are your organisation’s specification standards for AI work delegation? No organisational standards for how employees specify work for AI tools or agents Each person writes their own AI instructions, with no shared templates or quality standards Shared instruction templates exist for common tasks, with basic quality checks on what AI produces Step-by-step guides built into how teams work, with reusable templates, set quality checks, and documented good practice for handing work to AI An organisation-wide system with shared, version-controlled instruction templates, common standards across teams, ongoing improvement, and measured quality for AI-done workD15. GenAI & Agent Deployment This is about how far your organisation can roll out and manage generative AI and AI agents at scale — with the right access controls, monitoring, and oversight for everyday production use.How far has your organisation rolled out AI chatbots and AI agents for real use? No ability to roll out AI chatbots or agents Using AI chatbots only through public websites; no company-run AI agents Company AI chatbots given to staff, or AI agents set up with proper limits and approved connections Custom AI tools (like company chatbots, or assistants that answer from your own documents), or company AI agents with access set by role and their activity monitored AI chatbots and agents used widely in real operations — several working together, some running on a schedule on their own — all under proper controls, full records, and monitoringPreviousNextYou’re done — that’s all 15. Click Submit and we’ll instantly show you: Your overall organisational AI Readiness score and level — from AI Unaware to AI Catalyst A radar chart of your five pillars, so you see strengths and gaps at a glance Your strongest pillar — and the one to strengthen first A recommended set of AI uses cases your organisation can embark on based on your profile and industry. Remember: Level 2 — AI Ready is the recommended organisational success target for most organisation today. You don’t need an AI Centre of Excellence to win with AI — you need leadership commitment, clean ethics, a handful of real use cases, usable data, and the infrastructure to ship. Reassess every 6 months — organisational readiness moves faster than you’d expect. Previous Submit Form
