# AI Readiness Assessment Framework: Artebello’s Six-Pillar Methodology

An AI readiness assessment framework turns a broad question - “are we ready for AI?” - into evidence about strategy, people, workflows, data, technology, and governance. Artebello’s AI Readiness Index™ scores these six dimensions separately, then combines them into an organizational readiness score. The purpose is to choose a practical next step and expose dependencies before committing to implementation.

This is the public methodology for the current self-assessment. The score is a self-reported diagnostic, not an independent audit, compliance certification, or forecast of project success. A guided engagement can examine the evidence behind the answers and convert the results into an implementation plan.

## What the six pillars measure

| Pillar | Question for the leadership team | Useful supporting evidence |
| --- | --- | --- |
| Business | Do we know where AI creates value, who sponsors it, and what deserves investment? | Strategic priorities, budget owner, ranked use cases, outcome measures |
| People | Can our teams identify, deliver, and adopt appropriate AI uses? | Role-based skills assessment, training records, delivery capacity, staff feedback |
| Process | Do we understand the work well enough to change it safely? | Workflow maps, exception paths, process owners, baseline performance |
| Data | Is the relevant information trustworthy, accessible, and connected? | Quality checks, access controls, ownership, lineage, integration inventory |
| Technology | Can the organization run and integrate AI securely in production? | Architecture review, integration tests, identity controls, support arrangements |
| Governance | Are policies, accountability, and risk controls in place? | Approved-use policy, risk register, approval records, incident procedures |

Evidence matters more than a confident answer. “We have an AI policy” should lead to a dated policy, an accountable owner, and evidence that staff understand it. “Our data is accurate” should lead to recent checks on the specific data needed for the first use case.

## Questions and answer scale

The base assessment contains 18 questions: three in each pillar. The five answer options range from “Not at all” to “Advanced / leading,” with values 1 to 5. Industry and region selections can add contextual questions; the exact tailored question count is shown before the questionnaire begins.

Choose the answer that describes current practice. Planned training, an unsigned policy, and a future integration do not demonstrate an established capability. If departments disagree, record the disagreement and investigate the evidence rather than averaging their confidence informally.

The current scoring engine averages the answered questions within each pillar. The guided questionnaire requires the questions on each step to be answered before continuing. A partial set of answers can produce an intermediate score in code, but it should not be treated as a completed organizational assessment.

## How the readiness calculator works

For each pillar:

`pillar score = round(((average answer − 1) ÷ 4) × 100)`

The overall index is the rounded arithmetic average of the six pillar scores. Each pillar has equal weight in the overall index, even when a tailored pillar contains more questions. The scale is 0–100; it is not the probability that an AI initiative will succeed.

For three answers of 4, 4, and 3, the average is 3.6667. Converting that average gives 66.6667, which rounds to a pillar score of 67. If every answer is 1, the score is 0. If every answer is 3, it is 50. If every answer is 5, it is 100.

For a complete, reproducible example, see the [sample AI readiness report](/resources/example-ai-readiness-report). It lists all 18 base answers and shows how the six scores produce an overall score of 49.

## The four maturity levels

| Score | Artebello maturity level | Practical focus |
| --- | --- | --- |
| 0–39 | Nascent | Establish ownership, a baseline, and a manageable starting use case |
| 40–59 | Emerging | Turn experiments into a coordinated plan and repair the largest gaps |
| 60–79 | Operational | Strengthen adoption, monitoring, data, and governance before expanding |
| 80–100 | Transformational | Maintain controls and capability as the organization and technology change |

A level is a summary, not permission to launch. An organization with a strong average can still have an unacceptable exposure in privacy, security, oversight, or data quality. The [AI project launch checklist](/resources/checklist-04-ai-project-launch) applies decision gates to a particular initiative.

## Benchmarking and interpretation

The current comparison values are illustrative reference points: Business 54, People 46, Process 50, Data 44, Technology 56, and Governance 41. Their rounded overall average is 49. They are not estimated from a published, representative dataset of organizations in your industry.

The existing percentile-style calculation is also illustrative. It cannot establish how many real organizations scored above or below you. Use the comparison to understand how the diagnostic is presented, and use documented evidence and repeat assessments to judge progress. Comparing results across organizations with different profiles or question sets requires care.

A separate compliance-readiness result can be attached when the selected profile has applicable frameworks. That result does not add a seventh pillar or change the six-pillar overall score. Self-reported compliance readiness is not a legal determination or certification.

## Turning a score into decisions

1. Review the weakest pillars with their owners and inspect the supporting evidence.
2. Identify which gaps prevent the first proposed use case from being delivered safely.
3. Give each action an owner, a deadline, and a measurable acceptance criterion.
4. Choose a bounded pilot whose outcome can be compared with a baseline.
5. Reassess using the same profile after the organization has implemented the actions.

The diagnostic currently identifies up to three priorities by comparing pillar scores with the illustrative references. Leaders should still override that ordering when a mandatory control, a serious risk, or a concrete dependency needs attention first. The [first-90-days guide](/resources/guide-01-first-90-days) explains how to organize the follow-through.

## Common framework questions

### What are the three pillars of AI readiness?

There is no single three-pillar model shared by every readiness assessment. A compressed model can combine dimensions that another model separates. Before comparing a three-pillar score with another assessment, identify the author, definitions, questions, and weighting. Artebello uses six pillars; Microsoft’s published assessment uses seven. The [assessment comparison guide](/resources/guide-ai-readiness-assessment-comparison) explains the differences in scope.

### What are the five stages of AI readiness?

Stage counts depend on the chosen model. A project’s delivery stages are also different from an organization’s maturity levels. Artebello’s current diagnostic uses the four scored levels above. Its consulting journey - Assess, Advise, Build, Scale - describes engagement activities, not a second scoring system. Do not relabel the levels just to match a different framework’s terminology.

### Can this framework replace an AI governance review?

The assessment can reveal a need for one. A governance review examines actual policies, decisions, risk controls, vendors, and operating evidence. Use the [executive governance checklist](/resources/checklist-05-executive-review) to structure that review.

## Public guide and advanced support

This methodology and its worked examples are open to read and share. A guided assessment, facilitated evidence review, and organization-specific roadmap are consulting engagements that can be scoped with Artebello. [Take the free assessment](/assessment), or [discuss a guided review](https://calendly.com/artebello/30min).

## Sources and related frameworks

- [Microsoft AI Readiness Assessment](https://learn.microsoft.com/en-us/assessments/94f1c697-9ba7-4d47-ad83-7c6bd94b1505/) - a separately defined organizational assessment.
- [NIST AI Risk Management Framework](https://www.nist.gov/itl/ai-risk-management-framework) - risk management guidance, rather than an equivalent Artebello score.
- [Download the methodology summary](/downloads/ai-readiness-methodology.md).

Methodology reviewed against Artebello’s scoring implementation on 14 October 2026.