Domain 4 of 4 · Chapter 1 of 4

Assess AI business readiness and maturity

Ready for what? Starting from a named use case

Suppose a regional insurer's chief executive asks whether the company is ready for AI: AWS describes its generative AI workload assessment[1] as a method for evaluating and improving an organization's preparedness to create or update generative AI workloads. A plain yes or no would mislead, because readiness is always readiness for something: having the data, people, decision rules, and operational support that one intended business outcome needs. The same guide is built for organizations that have not defined requirements from the outset: it helps them identify potential use cases, assess readiness, and define and refine use case goals and scope. A company with only broad ideas therefore uses the assessment to turn them into measurable, feasible use cases; it does not need a finished specification before it starts.

Naming the use case changes the answer. An internal assistant that summarizes the insurer's own policy documents for staff needs a document collection, usage guidelines, and a team to support it. A system that recommends claim decisions affecting customers may need different data, human oversight, regulatory review, and production support. Being ready for the first says little about readiness for the second. The AWS Cloud Adoption Framework for AI (a whitepaper AWS now marks as historical reference) makes the same point from the data side: identify early which data assets and sources each initiative relies on, and work backwards from your opportunities towards data requirements[2]. A vendor demonstration does not settle the question either: it shows what the vendor's system does in the demonstrated setting, not whether your organization meets the prerequisites.

This topic's four-step diagnosis

This topic organizes the diagnosis an AI business strategist can run once a use case is named into four teaching steps, one for each skill the AWS AIB-C01 exam guide lists under Task 4.1: Assess AI business readiness and maturity[3]:

  1. Assess readiness across the six perspectives described in the next section.
  2. Rate maturity by placing each perspective on the four levels of the AWS generative AI maturity model.
  3. Diagnose gaps by naming the capabilities that stand between today and what the strategy needs.
  4. Prioritize investments by deciding which capability development comes first, recorded as a progression roadmap with owners and milestones.

Steps 1 and 2 measure different things: readiness is judged against one named use case, while a maturity level represents an organization's progress toward using generative AI capabilities[4].

AWS change guidance recommends reassessing readiness at a key milestone such as a pilot application deployment[5], so the roadmap follows the evidence instead of the first snapshot; the figure below traces the four steps and that loop.

This topic diagnoses and prioritizes; the exam guide domain[3] lists neighboring work as separate tasks: Task 4.2, "Establish data and infrastructure foundations for AI"; Task 4.3, "Lead enterprise-wide change and build AI-ready workforce capabilities" (Lead AI change and build workforce capabilities); and Task 4.4, "Scale AI from pilots to enterprise-wide deployments".

The takeaway: do not answer "are we ready?" until you can say "ready for which outcome?"

Named use case1. Assessreadiness2. Ratematurity3. Diagnosegaps4. PrioritizeinvestmentsProgression roadmapReassess atkey milestones
AIB-C01 Task 4.1 as a flow from a named use case: assess readiness, rate maturity, diagnose gaps, prioritize investments, reassess at key milestones

Six perspectives that define AI readiness

Readiness belongs to the whole organization: AWS notes that the successful adoption of generative AI requires a holistic understanding of multiple organizational dimensions[6], so business, people, and oversight capabilities are assessed alongside the technology.

The AWS Cloud Adoption Framework (AWS CAF) groups the organizational capabilities that adoption depends on into six perspectives, and AWS CAF for AI[7] applies the same six (business, people, governance, platform, security, and operations) to AI adoption. It defines a capability as an organizational ability to use processes to deploy resources, such as people and technology, to achieve an outcome. The AWS generative AI maturity model[6] evaluates each maturity level across the same six areas, calls them pillars of adoption rather than perspectives, and states that they align with and extend the AWS CAF. The pillars are one of the model's four aspects of generative AI maturity[6], alongside focus areas, key activities, and transformation strategy.

The AWS Well-Architected Framework also has six pillars[7], but those are architectural best-practice areas for designing and operating workloads (reliable, secure, efficient, cost-effective, and sustainable systems), not organizational adoption capabilities.

The six AWS CAF perspectives applied to an AI use case

Perspective What it covers Evidence a ready organization can show for the named use case
Business Strategic alignment and measurable impact on business goals The use case is mapped to a business objective; a business case and a funding commitment from key stakeholders exist
People Talent, skills, leadership, culture, and cross-functional collaboration Active sponsorship from business and technology leaders; a readiness baseline for the affected groups; AI literacy in the teams involved
Governance Risk management, compliance, and ethical guidelines Guidelines for responsible AI use; a known approval path and accountable decision-makers
Platform Infrastructure, tools, and data foundations Data sources cataloged and their quality evaluated for this use case; infrastructure and integration that can support it
Security Protection of data, privacy, and models Data access controls and a plan for protecting personal data the use case touches
Operations Lifecycle management, monitoring, and feedback Plans for performance monitoring, versioning, and incident response

The coverage column follows the maturity model's pillar descriptions[6], with leadership placed under People because AWS CAF People guidance asks for active and visible executive sponsorship from both technology and business functions[8]. The evidence column draws on the model's Envision-level activities for each pillar[9], which include identifying high-value data sources for selected use cases (Business) and evaluating data infrastructure and quality to determine feasibility (Platform), and on its Experiment-level security activities[10], which call for data access controls and a strategy for protecting personally identifiable information (PII).

Familiar readiness dimensions sit inside this model rather than beside it: leadership alignment and cultural preparedness are People capabilities, data quality and technical infrastructure are Platform capabilities, and governance frameworks belong to Governance.

Turning the model into questions

AWS's generative AI workload assessment questionnaire[11] shows what readiness questions look like in practice. Among others, it asks whether a high-level business case articulates objectives, benefits, and cost; whether key stakeholders such as the CFO, CIO or CTO, and COO have made a funding commitment; how the organization would describe its AI literacy and readiness to adopt generative AI; how it ensures compliance with data protection regulations; and how mature the existing systems are that would integrate with the new AI capability.

A cloud infrastructure inventory, however detailed, answers only part of the Platform question. Using it as the enterprise readiness score leaves out the organizational and oversight capabilities, and AWS notes that many organizations struggle to fully use AI because of weak technological foundations, resistance to change, and data quality issues[1]; an infrastructure inventory alone does not assess all of these barriers.

The takeaway: score the named use case against all six perspectives, and assess Platform alongside the other five perspectives, not as the whole assessment.

Measuring cultural and leadership readiness

This section covers a People question: will the people a use case affects actually change how they work? Its rule: measure willingness and ability directly, rather than relying on activity counts alone.

AWS organizational change guidance defines an organization readiness assessment[5] as a tool for understanding an organization's propensity, ability, and desire to adapt to change, typically conducted through a survey. The guidance was written for cloud transformation programs; the method carries over to AI adoption because it measures how people respond to a change, not the technology behind it. Its sample statements are rated on a 4-point Likert scale (strongly agree, agree, disagree, strongly disagree) and cover four areas:

Sample organization readiness survey statements (AWS change guidance)

Area Sample statement
Leadership Senior management actively supports the transformation.
Culture Mistakes will be treated as opportunities to learn instead of being punished as failures.
Training I have learned the necessary skills to work effectively in the new environment.
Communications A clear vision has been developed and communicated to employees.

The primary output is a set of baseline readiness scores across the surveyed dimensions, together with prioritized mitigation plans that feed the change strategy.

Surveys and interviews do different jobs

The survey sets a baseline across the workforce. Interviews expose where leaders diverge: AWS change guidance recommends interviewing key stakeholders with the same set of 7-10 questions[12] to establish a baseline of alignment and show where the change team needs to focus, and it observes that leaders are rarely on the same page. The same guidance also describes a stakeholder assessment, which identifies and captures information about the people who will be affected by the program[12] and highlights perception gaps between executive leadership, program leadership, and implementation teams.

Compare results across affected groups

An average hides differences. The readiness guidance advises carefully considering which demographic data to gather[5] before deploying the survey, so results can be compared across the groups the change touches. Strong enthusiasm at headquarters can coexist with a low score in a regional operations team whose daily work the use case changes most, and a segmented comparison can help reveal it.

What participation counts alone do not establish

An organization readiness assessment exists to understand the organization's propensity, ability, and desire to adapt to change[5], and training attendance records participation, not that desire. A person can complete every course and still not want to change how they work, so attendance alone does not establish cultural acceptance.

The takeaway: survey for the workforce baseline, interview for leadership alignment, and segment both by affected group.

The four maturity levels and the evidence for each

Step 2, rate maturity, uses the AWS generative AI maturity model[4], which is structured across four levels, each representing an organization's progress toward using generative AI capabilities. Its rule: place an organization by what it has demonstrably done, not by what it intends or has bought. Each level comes with criteria, the observable evidence that places an organization there. Three level names also name phases of the AWS CAF for AI adoption cycle, Envision, Align, Launch, and Scale[2], so keep levels and phases apart: a maturity level describes the state one perspective has demonstrably reached; a phase describes the work an initiative is doing now.

Level 1: Envision

Organizations explore generative AI concepts, build awareness, and identify potential use cases. The Envision criteria[9] are demonstrated basic knowledge of generative AI fundamentals, documented awareness of industry applications, an emerging understanding of cultural readiness, an initial exploration of use cases and benefits, and preliminary consideration of governance and security. The model's overview table adds a telling marker for this level: no formal projects or resource allocation.

Level 2: Experiment

Organizations validate generative AI's potential through proofs of concept and pilot projects. A proof of concept (PoC) is a small build that tests whether an idea is feasible and valuable; pilot projects are carefully controlled trials that test hypotheses about generative AI's value while minimizing risk[10]. The Experiment criteria are active pilots and PoCs in progress, dedicated cross-functional teams assigned to generative AI initiatives, a structured internal training program, selected and validated AI models and tools, and defined initial governance and data frameworks.

Level 3: Launch

Organizations move a select few proven solutions into production. The Launch criteria[13] are production-ready generative AI solutions delivering measurable business outcomes; baseline security, governance, and responsible AI frameworks; operational controls that include automated monitoring and alerting; a defined human-in-the-loop process (a step in which a person reviews or decides on the AI's output) for AI decisions; and preliminary roles and operational responsibilities for cross-functional AI teams.

Level 4: Scale

Organizations establish enterprise-wide capabilities. The Scale criteria[14] are widespread use of generative AI across multiple departments; an enterprise-wide AI infrastructure and tooling ecosystem; a defined and implemented operating model (how AI work is organized, owned, and supported across the company) and RACI matrix (a chart recording who is responsible, accountable, consulted, and informed for each activity); a library of standardized, reusable AI components that self-service capabilities make accessible across the organization; automated governance at enterprise scale; and evidence of sustained innovation practices and outcomes.

Two maturity boundaries

Envision versus Experiment is the line between awareness and structured doing. An organization whose executives intend to adopt AI, or that has bought many AI tool subscriptions, is still at Envision until pilots are active with assigned cross-functional teams, initial governance, and internal capability building. Purchases are not structured adoption practices.

Launch versus Scale is the line between a few supported applications and a repeatable capability. In AWS's words, Level 3 focuses on select production workloads, while Level 4 enables rapid deployment of a large number of generative AI applications through standardized and reusable components[14]. A single customer chatbot used by millions of people is still one deployment; Scale is about how many business units can build and adopt AI on shared, self-service foundations.

Maturity levels are not transformation phases

The AWS AIB-C01 exam guide also uses the level names for phases: its Task 4.4 skill is to apply iterative transformation approaches that progress through phases (for example, envision, experiment, launch, scale)[3]; see Scale AI from pilots to enterprise-wide deployments. With the level-versus-phase distinction from the start of this section, a company rated Launch for its Platform perspective can still take a brand-new initiative through an envisioning phase.

Because each maturity level is evaluated across the pillars of adoption[6], a business unit or use case placed on one level still has each perspective checked against that level's criteria, as this topic's maturity decision tree shows. The takeaway: name the level by its criteria, and test every claimed level against the evidence it requires.

Rating maturity per perspective and business unit

Do not rely solely on a company-wide label: each aspect of generative AI adoption should be considered independently[4]. AWS's wording says "aspect"; this topic applies the rule to each perspective and business unit, so rate each perspective on its own: the level one perspective reaches is not necessarily the level of another. AWS also notes that many organizations find characteristics of several levels across their teams and use cases, and that no single level is inherently superior or inferior: the appropriate level depends on the organization's goals and readiness.

Characteristics of multiple maturity levels in a hypothetical retailer

Perspective Customer service: level, AWS activity Finance: level, AWS activity
Business Launch: value dashboards Envision: use-case mapping
People Experiment: structured training Envision: internal training
Governance Experiment: model evaluation criteria Envision: preliminary governance consideration[9]
Platform Launch: automated CI/CD (continuous integration and continuous delivery) pipelines Envision: data infrastructure and quality evaluation[9]
Security Launch: security controls Envision: preliminary security consideration[9]
Operations Launch: SLA (service-level agreement) frameworks Envision: monitoring plans
  1. One label would hide the variation in this profile. Reporting this hypothetical retailer as "Launch" because customer service runs a production assistant hides that finance shows the Envision marker of no formal projects or resource allocation[4]. Using the most advanced department's rating to plan investment risks underfunding the foundations the weaker unit needs.
  2. A lagging perspective inside an advanced unit is itself a finding. Here, customer service runs production applications, yet its Governance perspective shows only Experiment activities such as model evaluation criteria[10]. The Launch criteria call for baseline security, governance, and responsible AI frameworks for production use, so that mismatch is a gap to diagnose in step 3.
  3. A higher level is not automatically the target. If finance's goal is a few internal forecasting aids, it may never need Scale; its goals and readiness set its progression, not the top of the model.

The takeaway: report maturity as a profile across perspectives and business units, and set target levels from business goals, not the model's ceiling.

Gap analysis: current state, future state, and what continues

Step 3, diagnose gaps, rests on one rule: a gap is the difference between the current state and the desired future state, and an honest gap analysis records what must continue as well as what changes.

AWS change guidance calls the tool a change impact assessment[15]. It reports how a change affects skills, processes, performance management, and technology for each stakeholder group, and it includes a gap analysis that documents the change between the current and future states. The guidance stresses that, in addition to identifying the changes, it is also important to document what stays the same. Its template columns make a practical checklist:

Change impact assessment template columns (AWS change guidance)

Column Question it answers
Impacted area Which area changes, for example leadership, culture, process, policies, structure, skills and capabilities, performance management, or systems?
Current state What happens today?
Future environment What is the desired future state?
Change gap or impact What are the major changes, what stays the same, and what has to continue?
Who is impacted? Which stakeholders or groups are affected?
Level of impact Is the impact high, medium, or low?
Change issues, risks, barriers What could prevent the change from succeeding?

Why "what continues" matters

For example, suppose a claims-triage assistant's future state has the AI draft a triage recommendation that an adjuster reviews: its gap analysis must still document what stays the same[15], not only what changes. A feature backlog limited to summarization and routing features could miss what must continue in this illustration: the insurer's response-time commitments to claimants, the adjuster's accountability for the final decision, and the follow-up questions the call center still handles. Those continuing obligations shape the real transformation gap, and they reveal affected stakeholder groups outside the project team. The guidance notes that the assessment helps identify stakeholders who are only tangentially tied to the change[15], and that its outputs feed communication plans, training plans, stakeholder engagement plans, and updates to the business case and risk log.

The takeaway: compare current and future states area by area, write down what continues, and treat a feature list as an input to the gap analysis, never as the gap analysis itself.

Capability gaps across people, process, technology, and governance

The AWS AIB-C01 exam guide lists the skill to identify specific capability gaps across people, process, technology, and governance that prevent successful AI transformations[3]. This topic groups capability gaps into those four plain teaching types. In this topic they are simplified teaching groupings of the six perspectives from earlier, not a second model or an exclusive partition of capabilities. Its rule: match the remedy to the gap type, because a remedy aimed at the wrong type spends money without moving the blocker.

An illustrative teaching grouping of gap types, related perspectives, and fitting remedies

Gap type Illustrative related perspectives Typical symptom Remedy that fits Remedy that does not fit
People People Nobody connects AI opportunities to products and customers; specialist skills are missing Develop, hire, or retrain for business-side AI roles and specialist roles Buying more AI tools
Process Business, Operations, Governance No defined path from pilot to production; no monitoring or incident process Define the process, its owners, and the handover Hiring more data scientists
Technology Platform, Security Technical access to approved data is missing; no safe environment for experimentation; no access controls Platform, data foundation, and security work A communication campaign
Governance Governance Nobody can say who approves a use case or how ethical and regulatory concerns are reviewed Define decision rights, review procedures, and accountable approvers A shared project tracker

People gaps include business talent

AWS CAF for AI names two kinds of talent that AI adoption needs[7]: technical talents (such as data scientists, applied scientists, and machine learning engineers) and non-technical product talents (such as ML product managers and ML strategists) who manage roadmaps and identify needs. It also recommends complementing highly specialized scientists with business-close counterparts. So compare the roles and skills the strategy requires with the current workforce, then separate capabilities you can develop internally from specialist gaps that need hiring or partners (the next section compares retraining, hiring, and partnering). A larger data science hiring target does not fix weak business ownership: technical headcount does not supply product or domain accountability.

Approval gaps require authority, not just coordination

The AWS Responsible AI Lens asks teams to check the organization's policies and processes for approving AI use cases[16], which may set out approval requirements and designate responsible oversight bodies, and to determine which teams or committees need to sign off before the project can proceed; it rates the risk of skipping this practice as High. When a team cannot name who may approve its use case for go-live, or how ethical and regulatory concerns get reviewed, the missing capability is governance, even if delivery is well coordinated. Visibility is not authority: a shared project tracker shows everyone the status but gives nobody the right to decide. Assess whether the organization can make and evidence accountable decisions separately from whether it can execute project tasks. The AWS AIB-C01 exam guide lists designing those structures as a separate task, Task 3.2: Establish AI governance structures and ensure regulatory compliance[17]; see Establish AI Governance and Regulatory Compliance.

The takeaway: name the gap type first, then choose a remedy of the same type.

Closing skill gaps: retrain, hire, or partner

This section covers a sourcing decision that the gap analysis forces often: how to obtain missing expertise. Its recommendation: partners can bridge specialist gaps for bounded work with knowledge transfer, while AI strategy leadership should normally stay in-house.

This topic compares three complementary approaches:

  1. Retrain existing staff when deep business knowledge matters most. AWS CAF for AI notes that transitioning some existing talent to AI roles is beneficial for organization-wide adoption[7] and that continuous retraining produces people who have in-depth business knowledge and can run projects.
  2. Hire when the strategy needs specialist depth the organization will need for the long term. The same guidance advises aligning the hiring strategy tightly with the overall AI strategy and ambition.
  3. Partner when expertise is needed now and cannot be supplied in time. The Experiment level's People activities include partnering with external specialists, such as the AWS Generative AI Innovation Center or AWS Professional Services, to co-build a PoC, provide support, and transfer knowledge[10].

The approaches combine rather than compete: partner support can complement hiring and retraining.

The boundary on outsourcing

AWS CAF for AI states that AI strategy leaders need to be tightly connected to the business and drive value from within, a role that can seldom be handled by an outsourced firm[7]. For the skill-building engagement discussed here, give the partner a bounded scope and a knowledge-transfer plan, and retain internal ownership of the strategy.

AWS CAF People guidance does allow partners and managed service providers to temporarily or permanently augment your workforce[8]. Permanent augmentation can be a valid staffing choice; without learning or knowledge transfer, augmentation alone does not build internal capability. A permanent external delivery contract presented as the workforce development plan fails because delivery alone does not transfer capability to the organization.

The takeaway: retrain for business knowledge, hire for lasting specialist depth, partner for speed with knowledge transfer, and keep the strategy owner internal.

Prioritizing and sequencing development investments

Step 4 starts by deciding which capability investments come first, the first half of the exam guide's skill to prioritize development investments[3]. Its rule: rank by strategic impact and feasibility at the current maturity level, then let blocking dependencies set the order.

The exam guide lists ranking AI initiatives as a separate skill, to prioritize AI initiatives based on business value, feasibility, sustainability, and strategic alignment (by deciding to scale, pause, or terminate initiatives)[18]; see Align AI strategy with business goals. This section ranks the capability investments (data, skills, governance, platform) that those initiatives depend on. The criteria look alike, but they decide different things: an initiative can rank first on business value and still have to wait for a capability investment it depends on.

Rank by impact and feasibility together

The maturity model's Envision guidance recommends a use case matrix that prioritizes projects by feasibility, business impact, and alignment with strategic goals[9], and its Experiment guidance prioritizes strategic use cases based on business value and feasibility[10]. Apply the same pairing to capability development. A high-value opportunity with unmet prerequisites may need foundational work before a pilot is justified. A feasible project earns priority when it advances a meaningful business objective, not because it is easy to demonstrate or technically novel.

Unblock first

When a strategic initiative depends on unavailable data, unresolved data ownership, or missing workforce skills, fund the prerequisite that blocks responsible progress before work that depends on it. AWS CAF for AI describes foundational capabilities as what, above all else, drives success or failure when adopting AI[2], and AWS CAF Governance guidance on program management calls for managing interdependencies by aligning multiple initiatives[19]. Coordinate people, process, technology, and governance investments so that no workstream runs ahead of its dependencies. For example, extra compute funding does not resolve an initiative's dependency on approval to use customer data. Work that does not depend on the blocked capability, such as preparing the team that will use the system, can proceed in parallel.

Fund shared foundations once

When several strategic use cases need the same data foundation, skills, or governed access, identify the overlap before funding separate solutions. AWS maturity guidance for moving beyond Launch recommends assessing existing implementations to identify reusable patterns and components[13], and the Scale level drives adoption through reusable components and standardized patterns that deliver value[14]. Three departmental funding requests for "data cleanup" may describe one common prerequisite; independently rebuilding that same foundation risks duplicating the enabling investment. Keep a shared foundation tied to the initiatives it enables, so its business justification is assessed against expected outcomes.

Following AWS CAF guidance on managing interdependencies by aligning multiple initiatives[19], the figure below ranks one candidate by impact and feasibility before any funding branch and funds an unblocked candidate in ranked order.

The takeaway: impact and feasibility decide what is worth doing, and dependencies decide the order.

Rank by impactand feasibilityBlocked by a missingprerequisite?YesNoDo other initiativesneed it too?Fund inranked orderYesNoFund one sharedfoundationFund theprerequisite first
Ordering a capability investment: rank by impact and feasibility, then fund blocking prerequisites first and shared foundations once

Building the progression roadmap

The second half of step 4 creates what the exam guide calls progression pathways based on strategic objectives and the current business maturity level[3], recorded here as a progression roadmap. Its rule: a maturity label becomes useful only when it drives specific, owned development commitments.

The maturity model's transformation strategy guides organizations through incremental improvements, including developing a generative AI roadmap and a data strategy, aligning with business goals, investing in talent and tools, and implementing governance frameworks[6]; AWS presents the model's four aspects as a way to identify gaps, prioritize investments, and create actionable plans for progression. This topic recommends giving every roadmap improvement four things:

  • An owner accountable for delivering it.
  • Resources: budget and people.
  • Dependencies: which other items must be in place first (see Prioritizing and sequencing development investments).
  • A measurable milestone that shows when it is done.

AWS CAF Governance guidance adds that you regularly validate your roadmap with your business sponsors and escalate issues to senior leadership in a timely fashion[19] to drive accountability and transparency. A target maturity score such as "reach Launch by next year" is a goal, not a plan: it assigns and resources nothing.

What each transition typically funds

The maturity model names the moves that take an organization to the next level. Its transformation strategy guides the organization's maturity advancement[6], and the figure below labels each of the three transitions with its main investment.

  • Envision to Experiment: fund small proofs of concept. The Envision transformation strategy[9] is to form cross-functional generative AI squads with clear roles, build a use case matrix and a short list of top use cases for PoCs, secure budget and personnel for small-scale PoCs, upskill staff on tools such as Amazon Bedrock and Amazon SageMaker AI, complete preliminary governance, begin planning organizational change management, and define success criteria and metrics for each PoC.
  • Experiment to Launch: build production foundations. The Experiment transformation strategy[10] is to create production-grade infrastructure, implement production-grade governance and observability, focus on compliance, build a dedicated team that maintains standardized paths to production, and implement operational excellence (incident response and escalation, service-level agreements, and cost optimization).
  • Launch to Scale: make capabilities reusable. The Launch transformation strategy[13] is to formalize the operating model and RACI matrix, identify reusable patterns and components, expand use cases across departments, turn the platform into a self-service internal platform, share knowledge, scale operations, invest in analytics (for example, Amazon Quick for visualization) for continuous improvement, and check whether data governance supports self-service.

Early stage: buy learning before large commitments

For an organization moving beyond Envision, the first investment is bounded learning: secure budget and personnel for small-scale PoCs and define success criteria for each one, and, as the Experiment guidance puts it, limit the scope of PoCs to a manageable scale with clear success metrics[10]. This builds practical capability and tests value before a broad production program. An enterprise-wide license commitment made to win early stakeholder confidence gets the order backwards, because the spend comes before any validation of the use cases. AWS CAF for AI warns that trying to do everything in the same cycle can lead to discouragement in the organization[2], and recommends pairing the larger picture with many pragmatic, actionable steps and measurable key performance indicators (KPIs) on those smaller steps. Production-grade and enterprise-wide capabilities become appropriate investments as evidence and maturity support the next step.

The takeaway: turn each gap into an owned, resourced, dependency-aware milestone, and size early investments to buy evidence.

EnvisionFund smallproofs of conceptExperimentBuild productionfoundationsLaunchMake capabilitiesreusableScale
The three transitions between the four AWS generative AI maturity levels and the main investment each one funds

Reassessing readiness as conditions change

The loop from step 4 back to step 1 rests on one rule: a readiness assessment is a snapshot, so reassess at milestones and let the findings change the roadmap.

AWS change guidance recommends conducting an organization readiness assessment at a key milestone such as a pilot application deployment, and using it periodically to measure progress[5]; its best practices include planning follow-up assessments throughout the program. The same guidance lists what the findings can justify:

  • Reviewing the strategic vision and business case.
  • Obtaining additional sponsorship.
  • Expanding ownership to cross-functional leaders.
  • Investing in more communications and training.
  • Prioritizing skill building.

The stakeholder assessment (introduced in Measuring cultural and leadership readiness) should likewise be reviewed and updated to reflect changes in the project, scope, impacts, and stakeholder turnover[12], meaning people who leave and join.

Share the results with the people who took part. AWS guidance notes that sharing results with participants shows progression, empathy, and program velocity[12], and it also makes the basis for roadmap changes visible, so a revised schedule or a new training investment does not look arbitrary.

The original assessment is not a permanent approval. If the scope grows from an internal pilot to a customer-facing service, or the sponsoring executive leaves, the earlier findings no longer describe the situation, and the roadmap built on them needs revisiting.

The takeaway: schedule reassessments at milestones, publish the results to participants, and adjust sponsorship, schedule, communications, or training when the evidence changes.

Practice scenarios: readiness and maturity

In these practice scenarios, an organization has a symptom and several constructive-sounding options; look for the action that addresses the capability blocking the named outcome.

Scenario cue Suggested response Less suitable response, and why
"Is the company ready for AI?" with no use case named Name the target use case, then assess all six perspectives against its prerequisites Point to a successful vendor demonstration: it does not establish the company's own prerequisites
Readiness judged from an infrastructure review alone Extend the assessment to Business, People, Governance, Security, and Operations Use the infrastructure inventory as the readiness score: it omits organizational and oversight capabilities
High training completion, unclear acceptance Run a readiness survey and stakeholder interviews, compared across affected groups Treat attendance as cultural acceptance: participation is not willingness to change
Many AI subscriptions bought, no pilots running Gather evidence for each perspective before assigning a level Rate it at Experiment from purchases alone: purchases are not structured adoption practices
One production assistant with a very large audience Check the assessed capability against all Launch criteria: measurable outcomes, baseline security, governance and responsible AI frameworks, automated monitoring and alerting, a human-in-the-loop process, and defined cross-functional team roles and operational responsibilities Rate it at Scale from audience size alone: one deployment's audience is not enterprise-wide repeatability
One department far ahead of the others Rate perspectives and units separately and invest by the uneven profile Plan with the most advanced department's rating: it hides weaker capabilities
Team cannot say who approves go-live Governance gap: define decision rights and review procedures Add a shared project tracker: visibility does not confer decision rights
Weak business ownership of AI initiatives People gap in business-side roles such as product and domain accountability Raise the data science hiring target: technical headcount does not supply business ownership
Initiative stalled waiting for data-use approval Resolve the approval dependency first and continue independent work in parallel Add compute funding: capacity does not resolve an approval dependency
Several departments request separate budgets for the same data work Assess overlap and identify reusable patterns and components before choosing shared funding Fund separate implementations without checking reuse: this risks duplicating the enabling investment
Specialist skills missing, deadline fixed A bounded partner engagement with knowledge transfer while strategy leadership stays internal A permanent external delivery contract as the workforce plan: delivery alone transfers no capability
Leadership wants to "reach Launch this year" A roadmap with owners, resources, dependencies, and milestones Treat the target score as the plan: a score assigns and resources nothing
Organization just past awareness wants a visible commitment Fund bounded PoCs with explicit success criteria An enterprise-wide license commitment: the spend precedes validation
Scope or sponsor changed since the last assessment Reassess readiness and update the roadmap Rely on the original assessment as a permanent approval

In these practice scenarios, ask which capability is missing for the named outcome and consider a remedy that addresses it, including dependencies across perspectives (see Capability gaps across people, process, technology, and governance above). The takeaway: diagnose first, then choose the remedy that matches the diagnosis.

The four levels of the AWS generative AI maturity model

CategoryEnvisionExperimentLaunchScale
What the organization is doingBuilding awareness, exploring concepts, and identifying candidate use casesRunning structured proofs of concept and pilots to validate business valueRunning a select few production applications that deliver measurable outcomesAdopting AI widely across departments through reusable, self-service components
Evidence that places you hereBasic AI knowledge; no formal projects or resource allocation; an emerging view of cultural readinessActive pilots; dedicated cross-functional teams; a structured training program; initial governance and data frameworksBaseline security, governance, and responsible AI frameworks; automated monitoring and alerting; a defined human-in-the-loop processEnterprise-wide AI tooling; an implemented operating model and RACI (responsible, accountable, consulted, informed) matrix; automated governance at enterprise scale
Investments that move you to the next levelCross-functional teams, a use case prioritization matrix, budget and people for small proofs of concept, preliminary governance, success metrics per proof of conceptProduction-grade infrastructure and governance, observability, compliance, a dedicated AI team, incident response and service-level agreementsA formal operating model, reusable components, a self-service platform, knowledge sharing, data governance that supports self-serviceHighest level: keep identifying high-value use cases and improving through feedback and monitoring
Common misclassificationCounting purchased AI subscriptions as ExperimentCounting a successful demonstration as readiness for productionCounting one large deployment's audience as ScaleAssuming Scale is the right target for every use case

Decision tree

Widespread adoptionacross departments onreusable components?YesScaleNoA few productionapplications deliveringmeasurable outcomes?YesLaunchNoActive proofs of conceptor pilots run byassigned teams?YesExperimentNoEnvisionPlace one business unit or use case,then check every perspectiveagainst that level's criteria

Sharp facts the exam loves — give these one last read before exam day.

Cheat sheet

Sharp facts the exam loves — scan these before test day.

AI readiness spans business capabilities as well as technology

AWS CAF-aligned AI adoption guidance assesses business, people, governance, platform, security, and operations together. Business addresses strategic outcomes, people addresses skills and collaboration, and governance addresses risk and decision rules; platform, security, and operations address enabling technology, protection, and reliable service delivery. Use these perspectives to locate the capability that needs attention rather than treating technical sophistication as a complete readiness assessment.

Trap Using a cloud infrastructure inventory as the enterprise readiness score because it omits organizational and oversight capabilities.

5 questions test this
Readiness must be assessed against a specific business use case

An AI readiness assessment should connect the intended business outcome to the data, organizational capabilities, and operational requirements needed to achieve it. A company with only broad ideas can use the assessment to refine them into measurable, feasible use cases. Being ready for a bounded internal assistant does not establish readiness for a consequential customer-facing decision process with different dependencies.

Trap Using a vendor demonstration's success as readiness evidence because it does not establish the company's own prerequisites.

5 questions test this
Organizational readiness includes willingness and ability to change

An organizational readiness assessment examines people's desire and ability to adopt new ways of working, alongside cultural and organizational barriers. Surveys can establish a readiness baseline, while stakeholder interviews expose differences in leadership understanding and commitment. Compare responses across affected groups so enthusiasm at headquarters does not conceal a workforce readiness gap elsewhere.

Trap Using training attendance as evidence of cultural acceptance because participation does not establish willingness to change.

5 questions test this
Awareness and structured experimentation represent different maturity levels

The AWS generative AI maturity model calls foundational awareness and use-case exploration Envision, and active, structured proofs of concept and pilots Experiment. Experiment includes assigned cross-functional teams, initial governance, and internal capability development. Classify an enterprise by demonstrated activities and supporting practices, not by executives' intentions or the availability of an AI tool.

Trap Using the number of purchased AI subscriptions to assign maturity because purchases do not establish structured adoption practices.

7 questions test this
A few supported production applications do not yet constitute enterprise-scale maturity

In the AWS generative AI maturity model, Launch means a select set of production applications delivers measurable outcomes with governance, monitoring, and support. Scale adds widespread departmental adoption, reusable capabilities, self-service access, and an implemented enterprise operating model. A reliable standalone deployment demonstrates a different level of maturity from a repeatable capability that many business units can use.

Trap Using the size of a single deployment's audience as the maturity measure because it does not establish enterprise-wide repeatability.

6 questions test this
AI maturity can differ across capabilities and business units

The AWS generative AI maturity model treats maturity as a continuum within each adoption aspect, rather than requiring every team and capability to share one label. Assess uneven strengths and gaps separately, and select the appropriate progression in the context of business goals and readiness. A more advanced level is not automatically the right immediate target for every use case.

Trap Using the most advanced department's maturity rating for investment planning because it hides weaker capabilities elsewhere.

6 questions test this
An AI workforce gap can involve business expertise as well as technical skills

AI transformation needs both technical specialists and people who connect AI opportunities to products, processes, and customer value. Compare the roles and skills required by the strategy with the current workforce, then distinguish capabilities that can be developed internally from specialist gaps that require hiring or partners. Existing employees with deep business knowledge can be valuable candidates for transition into AI-related roles.

Trap Using a larger data science hiring target to solve weak business ownership because technical headcount does not supply product or domain accountability.

6 questions test this
Gap analysis must distinguish what changes from what must continue

Compare the current and desired future states of AI-affected processes, roles, policies, skills, and technology. Record both the capabilities that must change and the responsibilities or service obligations that remain necessary. This establishes the actual transformation gap and identifies affected stakeholder groups, including those outside the immediate project team.

Trap Using the new application's feature backlog as the complete gap assessment because features do not capture unchanged obligations or organizational impacts.

5 questions test this
Missing approval authority is a governance capability gap

AI governance readiness requires defined authority, risk review expectations, and accountability for decisions. When teams cannot identify who may approve a use case or how ethical and regulatory concerns are reviewed, the missing capability is governance even if delivery activities are well coordinated. Assess whether the organization can make and evidence accountable decisions, separately from whether it can execute project tasks.

Trap Using a shared project tracker as the remedy for missing go-live authority because visibility does not confer decision rights.

6 questions test this
AI development priorities should combine strategic impact and feasibility

When choosing which capabilities and use cases to develop first, compare expected business impact, strategic alignment, and feasibility at the current maturity level. A high-value opportunity with unmet prerequisites may need foundational work before a pilot is justified. A feasible project earns priority when it advances a meaningful business objective, not merely because it is easy to demonstrate.

Trap Using technical novelty to rank investments because novelty does not establish business impact or organizational feasibility.

6 questions test this
Blocking capability gaps determine investment sequence

When a strategic AI initiative depends on unavailable data, unresolved ownership, or missing workforce capabilities, prioritize the prerequisite that prevents responsible progress. Coordinate investments across people, process, technology, and governance so one workstream does not advance beyond its dependencies. Other improvements can proceed in parallel when they do not depend on the blocked capability.

Trap Using additional compute funding to accelerate an initiative awaiting data-use approval because capacity does not resolve that dependency.

6 questions test this
Shared capability needs can justify investments that benefit several initiatives

When several strategic use cases need the same data foundation, skills, or governed access capability, identify the overlap before funding separate solutions. A shared investment can remove repeated barriers and establish a reusable foundation for later initiatives. Keep its justification tied to the business initiatives it enables rather than building a platform without identified consumers.

Trap Using separate departmental funding requests to treat a common prerequisite as unrelated local work because this duplicates the enabling investment.

6 questions test this
Partners can bridge skill gaps while internal leaders retain the business strategy

When an AI initiative needs expertise that the organization cannot supply immediately, use external specialists to accelerate bounded work and transfer knowledge to employees. Keep strategy leadership closely connected to the business and develop the internal capabilities needed to sustain the initiative. Partner support can complement hiring and retraining rather than replacing all internal ownership.

Trap Using a permanent external delivery contract as the workforce development plan because delivery alone does not transfer organizational capability.

6 questions test this

An AI progression roadmap connects the current capability state to the desired business outcomes through prioritized improvements in talent, tools, data, and governance. Give the work owners, resources, dependencies, and measurable milestones so progress can be evaluated. A maturity label becomes useful when it guides specific development commitments rather than remaining an assessment result.

Trap Using a target maturity score as the entire transformation plan because a score does not assign or resource the improvements.

6 questions test this
Early-stage investment should fund bounded learning before enterprise commitments

For an organization moving beyond awareness, secure budget and people for manageable proofs of concept with explicit success criteria. This develops practical capabilities and tests value before committing to a broad production program. Production-grade and enterprise-wide capabilities become appropriate investments as evidence and maturity support the next step.

Trap Using an enterprise-wide license commitment to secure early stakeholder confidence because the investment precedes validation of the use cases.

6 questions test this
Readiness findings should change the roadmap when organizational conditions change

Reassess organizational readiness at important milestones and update the progression plan as scope, stakeholder participation, and capability needs evolve. Findings may justify additional sponsorship, a revised schedule, or different communications and training investments. Sharing assessment results with participants also makes the basis for changes visible instead of leaving the roadmap disconnected from current readiness.

Trap Using the original readiness assessment as a permanent approval because it does not reflect later changes in scope or staffing.

6 questions test this

References

  1. https://docs.aws.eu/prescriptive-guidance/latest/gen-ai-workload-assessment/introduction.html
  2. https://docs.aws.amazon.com/whitepapers/latest/aws-caf-for-ai/your-ai-transformation-journey.html
  3. https://docs.aws.amazon.com/aws-certification/latest/ai-business-strategist-01/ai-business-strategist-01-domain4.html
  4. https://docs.aws.amazon.com/prescriptive-guidance/latest/strategy-gen-ai-maturity-model/overview-levels.html
  5. https://docs.aws.amazon.com/prescriptive-guidance/latest/oca-framework-align-leaders/readiness-assessment.html
  6. https://docs.aws.amazon.com/prescriptive-guidance/latest/strategy-gen-ai-maturity-model/overview-aspects.html
  7. https://docs.aws.amazon.com/pdfs/whitepapers/latest/aws-caf-for-ai/aws-caf-for-ai.pdf
  8. https://docs.aws.amazon.com/whitepapers/latest/overview-aws-cloud-adoption-framework/people-perspective.html
  9. https://docs.aws.amazon.com/prescriptive-guidance/latest/strategy-gen-ai-maturity-model/level-1.html
  10. https://docs.aws.amazon.com/prescriptive-guidance/latest/strategy-gen-ai-maturity-model/level-2.html
  11. https://docs.aws.eu/prescriptive-guidance/latest/gen-ai-workload-assessment/readiness.html
  12. https://docs.aws.amazon.com/prescriptive-guidance/latest/strategy-ocm/align-leaders.html
  13. https://docs.aws.amazon.com/prescriptive-guidance/latest/strategy-gen-ai-maturity-model/level-3.html
  14. https://docs.aws.amazon.com/prescriptive-guidance/latest/strategy-gen-ai-maturity-model/level-4.html
  15. https://docs.aws.amazon.com/prescriptive-guidance/latest/oca-framework-align-leaders/impact-assessment.html
  16. https://docs.aws.amazon.com/wellarchitected/latest/responsible-ai-lens/raiuc05-bp01.html
  17. https://docs.aws.amazon.com/aws-certification/latest/ai-business-strategist-01/ai-business-strategist-01-domain3.html
  18. https://docs.aws.amazon.com/aws-certification/latest/ai-business-strategist-01/ai-business-strategist-01-domain2.html
  19. https://docs.aws.amazon.com/whitepapers/latest/overview-aws-cloud-adoption-framework/governance-perspective.html