Study Guide · AIB-C01

AIB-C01 Cheat Sheet

382 entries · 13 chapters · 4 domains

AI Fundamentals and Literacy

Describe core AI concepts and terminology

Read full chapter

Cheat sheet

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

AI, machine learning, and generative AI describe related but different capabilities

Artificial intelligence is the broad category of systems performing tasks such as perception, reasoning, and prediction; machine learning is an approach within AI that learns patterns from data. Deep learning is a form of ML using multilayer neural networks, and modern generative AI uses learned patterns to produce content such as text, images, or audio. A system can classify transactions or recommend products without generating new content, while a generative model can also perform tasks such as classification when prompted.

Trap Using content generation as the defining requirement for every AI application, which excludes predictive and recognition capabilities.

4 questions test this
An algorithm is a method, while a trained model contains learned relationships

A learning algorithm is the procedure used to discover patterns, whereas a trained model is the resulting representation of those patterns and learned parameter values. The same learning procedure can produce different models when trained on different business data. When discussing an AI proposal, distinguish changing the learning method from updating the learned model with new training evidence.

5 questions test this
Training learns patterns, while inference applies a model to a new request

Training uses examples to learn model parameters; inference uses an already trained model to produce a prediction or response for an input. Forecasting next month's demand or generating a reply is inference, even when the input contains information the model has not previously encountered. Ordinary inference does not itself update the model's trained parameters; learning from newly collected examples requires an explicit training or updating process.

Trap Treating a growing archive of prediction requests as automatic retraining, which confuses collecting examples with updating model parameters.

5 questions test this
Foundation models provide broadly reusable capabilities rather than one narrowly trained task

A foundation model is pretrained on broad, diverse data and can be adapted or prompted for multiple downstream tasks. This allows a business to reuse an existing capability instead of funding training from scratch for every application. Broad pretraining does not establish that the model knows a company's private information or meets its specific quality requirements, so business-specific evaluation and, where necessary, adaptation remain important.

Trap Commissioning a new model from scratch for each text task, which overlooks the reusable capabilities of pretrained foundation models.

4 questions test this
Language, vision, and speech capabilities address different forms of business information

Natural language processing interprets or generates human language, supporting sentiment analysis, intent recognition, and translation. Computer vision extracts information from images or video, supporting visual inspection and object recognition. Speech recognition converts spoken words into text, which can then be analyzed for meaning; transcribing a call and determining customer sentiment are different tasks that can be combined.

Trap Using speech transcription alone to measure customer sentiment, which produces words rather than an interpretation of their meaning.

4 questions test this
A large language model generates language by predicting successive tokens

A large language model (LLM) is a model trained on extensive language data to learn patterns that support understanding and generating text. A generative LLM produces a response by predicting successive tokens from the supplied context and the tokens already generated. LLMs are language-focused members of the broader foundation-model category, which also includes models for non-language content such as images; foundation model and LLM are therefore not interchangeable names.

A learned prediction is an estimate rather than a guaranteed outcome

A learned prediction is a probabilistic estimate based on patterns, not a promise that the predicted event will occur. A demand forecast can therefore be useful even though actual demand differs from its predicted value. Generative responses can also vary across repeated requests with the same prompt because generation can select among possible next tokens; identical wording is not guaranteed simply because the request is unchanged.

4 questions test this
Broad task capability does not by itself establish artificial general intelligence

Artificial general intelligence (AGI) denotes the theoretical goal of generalized human-level cognitive ability, including learning and solving problems across domains beyond the tasks a system was developed for. Task-bounded AI operates within capability limits even when it performs many useful activities. Reusing a foundation model for several business tasks or letting an agent complete a workflow autonomously does not by itself demonstrate AGI.

Recommendation engines select relevant items rather than merely personalize wording

A recommendation engine uses evidence such as customer interactions, preferences, and item information to select or rank relevant products or content. Purchase history and product descriptions can help determine which existing items to show a customer. Generating a personalized message about an item is a different capability: it creates wording rather than performing the selection or ranking, although a business experience can combine both.

3 questions test this
Amazon Bedrock, Amazon SageMaker AI, and Amazon Quick serve different business capability needs

Amazon Bedrock provides managed access to foundation models and capabilities for creating generative AI applications and agents. Amazon SageMaker AI provides a managed environment for technical teams to develop, customize, train, deploy, and manage models across the custom-ML lifecycle. Amazon Quick is a ready-to-use business AI service combining assistance grounded in connected data, analytics, and workflow automation. These categories distinguish application-building capabilities, a model-development environment, and business-facing assistance; overlapping features do not make the offerings interchangeable.

4 questions test this
Supervised learning uses known outcomes to learn categorical or numerical predictions

Supervised learning learns from inputs paired with known target outcomes, such as historical transactions marked fraudulent or genuine. Classification predicts a category, including two-category and multiple-category outcomes; regression predicts a numerical value such as a selling price. A business must identify both the outcome it wants predicted and whether reliable historical examples of that outcome exist.

Trap Using customer segmentation as the training objective for an exact revenue forecast, which discovers groups rather than learning the required numerical target.

4 questions test this
Unsupervised learning discovers relationships without examples of a required answer

Unsupervised learning finds patterns in data without supplied target labels. Clustering groups similar customers or documents, while association discovery identifies relationships such as products frequently purchased together. These findings support exploration, but the resulting groups need business interpretation rather than being treated as predefined outcomes the system was taught to predict.

Trap Treating newly discovered customer clusters as verified churn labels, which substitutes similarity groups for observed business outcomes.

4 questions test this
Limited manual labels do not mean all learning must stop

Semi-supervised learning uses a small labeled dataset together with a larger unlabeled dataset, which can be useful when expert labeling is costly. Self-supervised learning derives learning targets from the data itself, as in learning to predict missing or subsequent text, and is widely used in foundation-model pretraining. These approaches differ from collecting a human-provided correct answer for every training example; neither removes the need to evaluate results against the intended business task.

Trap Requiring experts to label every document before any pretraining can begin, which overlooks learning signals obtainable from unlabeled data.

4 questions test this
Reinforcement learning learns actions from rewards rather than paired correct answers

Reinforcement learning improves a strategy through actions and feedback expressed as rewards, with the aim of improving cumulative outcomes over time. A locally attractive action can be less useful than a sequence that earns a better long-term result. Business leaders must understand what the reward encourages, because a system can learn to pursue the rewarded outcome even when that outcome is a poor proxy for the real business goal.

Trap Rewarding only immediate engagement for a long-term retention goal, which teaches the system to optimize a different outcome.

3 questions test this
Anomaly detection identifies unusual behavior rather than proving its cause

Anomaly detection identifies observations that differ from expected patterns, which can help surface unusual transactions or equipment behavior when exhaustive examples of failures are unavailable. An unusual observation is a signal for investigation, not by itself proof of fraud or a particular equipment fault. A known-outcome classifier instead learns the categories represented in its labeled examples, so the two capabilities answer different business questions.

Trap Treating an unusual transaction as a confirmed fraud label, which turns a deviation signal into an unverified conclusion.

3 questions test this
Data structure depends on its schema, not whether it contains numbers or words

Structured data follows a predefined schema, such as transaction rows with named fields for customer, date, and amount. Unstructured data includes narrative documents, images, and recordings whose meaning is not organized as fixed record fields. Short text fields can be structured, and a digital document can remain unstructured; the distinction helps determine whether a business needs record analysis or content understanding.

Trap Classifying all text as unstructured, which ignores schema-defined fields such as customer identifiers and product categories.

4 questions test this
Metadata organizes content without making all of its meaning structured

Semi-structured information combines organizational cues with content that does not fit a rigid table, such as a media collection with dates and topic tags. Those tags can support filtering and discovery, while understanding what a recording actually says still requires analysis of its content. A business should distinguish searching known metadata from extracting new meaning from the underlying material.

Trap Using document filenames as substitutes for analyzing contract obligations, which searches labels rather than the terms inside the documents.

4 questions test this
An AI solution must support the information modalities the business needs it to use

A modality is a type of information, such as text, images, audio, or video; a multimodal capability can process more than one type. A text-only capability cannot directly interpret visual evidence in a product photograph or a scanned chart merely because a user mentions it. For a mixed-information task, establish which modalities must be understood and whether any required conversion preserves the evidence needed for the decision.

Trap Using a text-only summary to judge visual damage that was never described in text, which removes the evidence needed for the assessment.

4 questions test this
Recognizing text and extracting business fields are different document-processing capabilities

Optical character recognition converts an image of text into machine-readable text. Intelligent document processing goes further by extracting, classifying, and validating business information such as invoice numbers, totals, and payment terms for downstream workflows. When the goal is to populate a business record rather than simply make a document searchable, assess field-level extraction and validation as well as text recognition.

Trap Accepting a searchable invoice as proof that its payment fields were correctly captured, which confuses text recognition with business-data extraction.

5 questions test this
Incorrect and missing records are data-quality problems before they are model problems

When predictions rely on inaccurate records or missing business information, prioritize correcting the source data and assessing its suitability before retraining. Accuracy means that recorded values reflect reality, while completeness concerns whether required information is present. Timeliness concerns currency and relevance, so an accurate old record may still be unsuitable for a decision that needs current information. Adding more examples from the same faulty process can enlarge the problem rather than supply better learning evidence.

Trap Funding a more sophisticated model for contradictory customer records, which leaves the incorrect learning evidence unchanged.

5 questions test this
Consistent definitions and valid values make data comparable across business sources

Data consistency requires equivalent business information to have compatible meanings and representations across records and sources. Validity checks whether values follow required formats or acceptable ranges; a value can pass a format check and still be factually wrong. When stores record the same business event differently, standardize collection definitions and quality checks rather than merely standardizing the AI's final reports.

Trap Standardizing forecast report layouts to reconcile incompatible store inputs, which changes presentation without making the underlying measurements comparable.

4 questions test this
Large historical datasets can still omit the people and situations that matter

Training evidence must represent the intended users and operating conditions, not simply contain many records. Underrepresented populations, rare outcomes, or missing business contexts can lead to uneven performance even when overall results look strong. Seek relevant coverage of those gaps and document exclusions; collecting more of the already dominant examples does not necessarily address them.

Trap Expanding only the majority customer segment to fix weak minority coverage, which increases volume without filling the missing experience.

4 questions test this
Training labels require quality control whether people or AI produced them

A label tells a supervised-learning system what answer to learn, so inconsistent or biased labels can teach the wrong associations. Require clear labeling guidance, checks against known examples, and review of disagreements by appropriate domain experts. Machine-generated labels also need validation against human judgment, especially for underrepresented groups and difficult cases; automation changes how labels are produced, not the need to establish their reliability.

Trap Treating agreement among automatically generated labels as independent proof of correctness, which can reproduce the same underlying labeling error.

4 questions test this
Synthetic examples are useful only when they preserve the business phenomena being learned

Synthetic or augmented data can supplement scarce examples, but its usefulness depends on realism, relevant relationships, and coverage of important edge cases. Require domain validation and checks for bias rather than accepting generated volume as evidence of quality. Document where synthetic examples can substitute for real observations and where their limitations make real data necessary.

Trap Replacing rare real incidents with generic generated examples, which may remove the unusual relationships the system needs to recognize.

4 questions test this
Data profiling reveals issues, but business expertise determines what they mean

Data profiling examines characteristics such as missing values, distributions, unusual observations, and inconsistent categories before training. Subject matter experts must interpret those findings because an unusual value may be either a recording error or a valid business event. Establish quality expectations and record the rationale for remediation instead of blindly accepting an automated suggestion or removing every unusual case.

Trap Removing every unusually large transaction before fraud training, which may discard legitimate examples of the events being investigated.

4 questions test this
Duplicate records can distort evidence even when every field is correct

Uniqueness is the data-quality dimension concerned with avoiding unintended duplicate records for the same business entity or event. Recording the same customer twice can inflate customer counts even when both records are accurate and complete. Assess whether repeated records represent genuinely separate observations before accepting record volume as evidence of the size of the business population.

Trap Using a required-field completeness check as the sole acceptance criterion for customer counts, because duplicate records can contain every required field.

Training, validation, and testing provide different kinds of evidence

Training data teaches the model, validation data supports development decisions, and a separate test set assesses performance after those decisions are made. A business approval should rely on examples not used to fit or repeatedly refine the model. The same independence principle applies to prompt engineering: use development cases to refine instructions and held-out cases to verify the final prompt. This distinguishes a reusable improvement from a result tailored to familiar examples.

Trap Using the final test set for every development revision, which makes the claimed independent assessment part of the development process.

3 questions test this
Information unavailable at decision time must not inflate historical evaluation results

Data leakage occurs when learning or evaluation uses information that would not legitimately be available for the real prediction. Examples include later business outcomes present in historical records and duplicate examples appearing across training and testing. When reviewing impressive results, ask whether the evidence reflects the information available at the actual decision point and whether evaluation examples are genuinely separate.

Trap Including completed-claim settlement information in an initial claim-risk prediction, which supplies knowledge the business would only obtain later.

2 questions test this
Training success must be compared with performance on unfamiliar examples

Overfitting occurs when a model learns training-specific detail or noise and performs well on training examples but poorly on new examples. Underfitting means the model has not learned the relevant relationship and performs poorly on both training and test examples. These patterns call for different technical investigations; business leaders should request evidence of generalization rather than assuming longer training or higher training accuracy is always better.

Trap Demanding more training solely to maximize an already strong training score, which can deepen memorization instead of improving generalization.

4 questions test this
Historical evaluation must address the specific context in which the model will be used

A national or historical average test score does not establish performance for a particular city, customer group, or newly changed market. Request evaluation on relevant subsets and, after a major business shift, suitably recent examples before relying on the broader result. The decision is whether the evaluation evidence supports the intended application, not whether a single aggregate score is impressive.

Trap Accepting a national average test result as sufficient evidence for entry into a poorly represented city, which leaves local performance unexamined.

3 questions test this
ISO/IEC 22989 and ISO/IEC 23053 support different levels of shared AI understanding

ISO/IEC 22989 establishes AI concepts and terminology for communication among diverse stakeholders. ISO/IEC 23053 provides a framework for describing the components and functions of AI systems using machine learning, including deep learning. Use shared vocabulary to resolve inconsistent terminology and the ML-system framework to explain how system elements relate; neither serves the same purpose as an organizational AI management system.

Trap Using ISO/IEC 42001 as the primary glossary for a terminology dispute, which selects a management-system standard rather than the dedicated vocabulary standard.

4 questions test this
ISO/IEC 42001 addresses organizational AI management rather than an individual model's accuracy

ISO/IEC 42001 specifies requirements for establishing, maintaining, and continually improving an AI management system. It applies to organizations that provide or use AI, not only those that develop models, and uses a Plan-Do-Check-Act approach to managing risks and opportunities. Its focus is coordinated policies, objectives, and processes across the organization rather than a substitute for evaluating an individual AI application's outputs.

Trap Using a model benchmark report as the organization's AI management system, which measures outputs without establishing management policies and improvement processes.

3 questions test this
The NIST AI RMF provides a voluntary structure for managing AI risks across the lifecycle

The NIST AI Risk Management Framework is voluntary, use-case agnostic guidance for incorporating trustworthiness into AI design, development, use, and evaluation. Its core functions are Govern, Map, Measure, and Manage: establish governance, understand context and risks, assess those risks, and prioritize responses. Govern is cross-cutting rather than a one-time first step; organizations tailor the risk-management outcomes and approaches to their application and context.

Trap Treating adoption of the NIST AI RMF as regulatory permission to deploy, which confuses voluntary risk guidance with applicable legal requirements.

3 questions test this
AWS CAF for AI treats adoption as an organizational capability challenge

AWS CAF for AI applies six perspectives to AI adoption: Business, People, Governance, Platform, Security, and Operations. These connect business outcomes and workforce capability with oversight, enabling technology, protection, and reliable operation. A technology assessment alone therefore cannot establish overall AI readiness; the perspectives provide a common way for business and technical stakeholders to discuss gaps without prescribing one executive owner for every organization.

Trap Using platform capacity as a complete AI-readiness assessment, which leaves people, governance, and other organizational capabilities unassessed.

3 questions test this
AWS responsible AI dimensions describe distinct concerns that a single quality score cannot replace

AWS describes responsible AI through fairness, explainability, privacy and security, safety, controllability, veracity and robustness, governance, and transparency. Explainability concerns understanding and evaluating outputs, while transparency enables informed stakeholder choices; controllability concerns mechanisms to monitor and steer behavior. Fairness considers impacts on different groups, and veracity and robustness address correct outputs even with unexpected inputs, so strong performance on one dimension does not establish all the others.

Trap Using a transparency disclosure as evidence that individual decisions are understandable, which substitutes stakeholder information for output explainability.

3 questions test this

Identify and select appropriate AI solutions

Read full chapter
  • Complete and stable rules favor deterministic automation
  • A workflow can combine AI interpretation with deterministic business controls
  • Automation choice follows an understood business process rather than a desire to use AI
  • An agent pursues a goal rather than merely providing a conversational answer
  • Agents can revise a plan by observing the results of intermediate actions
  • Tool use connects an agent's reasoning to current information and business actions
  • Agent memory preserves working context without necessarily changing the underlying model
  • Human oversight can require approval before action or provide visibility without blocking
  • Multiple agents are justified by distinct responsibilities, not by agent count alone
  • Orchestration defines who delegates, coordinates, and integrates agent work
  • Dependencies determine whether agents should work sequentially or in parallel
  • Agent-to-agent communication must preserve the information needed for the next responsibility
  • Correct individual agents do not guarantee a correct collaborative workflow
  • Ongoing monitoring needs a reference point and business-relevant response thresholds
  • Input changes and changed business relationships are different forms of drift
  • Prediction quality requires comparison with observed outcomes, not just healthy inputs
  • Retraining timing should follow evidence of need rather than an arbitrary calendar alone
  • An updated model must demonstrate improvement before replacing the current one
  • Monitoring delivers value only when findings feed an owned improvement process
  • A shared tool registry makes approval status visible when employees need it
  • AI-tool approval must identify the uses and data that the approval actually permits
  • Shadow AI can appear inside software the organization already approved
  • Practical approved alternatives reduce the incentive to use unsanctioned AI
  • Approval records and actual-use visibility answer different shadow-AI questions

Unlock with Premium — includes all practice exams and the complete study guide.

Apply generative AI concepts and techniques

Read full chapter
  • A prompt should name the exact task and any allowed answer choices
  • Relevant reference context supplies information that the task instructions alone cannot provide
  • Specify the output shape and limits when the business needs a particular deliverable
  • Audience and role context guide how an answer is framed rather than establish its truth
  • Source boundaries and an uncertainty response help keep answers within available evidence
  • Temperature influences response variation rather than factual correctness
  • Few-shot prompting demonstrates a desired pattern without fine-tuning the model
  • Prompt templates separate reusable instructions from changing business inputs
  • Complex requests can benefit from explicit intermediate tasks and dependencies
  • Useful refinement identifies a specific deficiency rather than asking for arbitrary regeneration
  • A tested prompt template needs reevaluation when its underlying model changes
  • The context budget includes more than the user's visible document
  • Missing input and a cut-off response point to different token constraints
  • A conversation's visible history is not automatically available in every model request
  • Long-content handling should preserve the evidence required by the particular task
  • Fitting more content into context does not guarantee that every detail will be used well
  • RAG supplies relevant external knowledge without retraining the answering model
  • A RAG answer can only be as well grounded as the evidence retrieved for it
  • RAG can provide source attribution, but citations still need to support the answer
  • Retrieval must respect access permissions before private information reaches the model
  • Semantic retrieval finds related evidence, while generation turns evidence into an answer
  • Fine-tuning adapts learned behavior rather than simply attaching reference documents
  • A fine-tuning proposal needs suitable examples and a justified training commitment
  • RAG and fine-tuning can address different gaps in the same business application
  • An adaptation succeeds only if evidence shows that it closes the intended business gap
  • Model distillation transfers demonstrated capability to a smaller model rather than refreshing its knowledge sources
  • Domain-language adaptation can learn from specialist text without paired ideal answers
  • Reinforcement fine-tuning adapts a model using feedback on response quality

Unlock with Premium — includes all practice exams and the complete study guide.

AI Strategy and Business Value Creation

Align AI strategy with business goals

Read full chapter

Cheat sheet

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

A useful AI strategy starts with a business problem and a measurable outcome

AI use-case discovery starts by identifying a customer or business problem, the decision or activity to improve, and the outcome that would justify investment. Map the required capability to that activity and define success before choosing a tool. Involve the business owner and affected stakeholders early so the proposal includes the process changes needed to realize the value.

Trap Using a technology demonstration as the investment brief, which proves a capability without establishing the business problem it should solve.

5 questions test this
Customer operations AI should address the actual service bottleneck

When customers need consistent answers to routine inquiries, use AI-assisted self-service grounded in approved information with a route to human support. When agents spend time searching, triaging, or documenting cases, agent assistance and summaries address that internal workload instead. Match the intervention to the bottleneck and assess resolution, recontacts, and customer experience rather than simply counting automated responses.

Trap Using a generic FAQ chatbot to resolve account-specific billing disputes, which lacks the account context and exception-handling authority that the work requires.

5 questions test this
Recommendations improve relevance while content generation improves creative throughput

When the business needs more relevant offers or product discovery, use personalization and recommendations based on customer preferences and context. When the bottleneck is producing campaign variants or product descriptions, use generative content assistance with brand review. The first intervention targets engagement, conversion, and upsell; the second targets production cycle time and the effort required to produce acceptable content.

Trap Using campaign draft volume as evidence that product recommendations are relevant, which measures content production rather than customer-item matching.

3 questions test this
AI can accelerate research by expanding the ideas and designs experts can evaluate

Research and development teams can use AI to generate design candidates, summarize extensive research, and explore alternatives under stated constraints. This creates value by accelerating ideation and narrowing the work experts must investigate, not by establishing that a proposed design is already valid. Keep domain validation in the process and judge the initiative by development time, useful candidates, and downstream quality.

Trap Using the number of generated prototypes as the release criterion, which measures exploration volume rather than validated product fitness.

5 questions test this
Software development AI creates value through delivery improvements rather than code volume

AI assistance can accelerate drafting code, creating tests, and understanding existing software, but the business benefit depends on the developer activity and surrounding workflow. Retain normal security, testing, and review requirements and evaluate time to delivery alongside quality and rework. An assistant that speeds one task may need workforce enablement and process changes before the team can deliver more useful software.

Trap Using generated lines of code as the productivity target, which rewards output quantity without showing usable software delivered.

5 questions test this
Knowledge assistance is valuable when employees struggle to find trusted company information

When employees repeatedly search policies, procedures, and product documents, use knowledge assistance over approved company content to shorten time to an answer. Identify which repositories contain the needed information and preserve users' access boundaries. Useful outcomes include faster onboarding, fewer unnecessary escalations, and less search effort; automating a transaction is a different business requirement from finding information.

Trap Using a general public chatbot as the company policy authority, which does not establish access to the organization's approved and permission-controlled information.

4 questions test this
Document extraction and classification address manual entry and document-handling bottlenecks

When employees must read forms or invoices and re-enter their contents, use AI-assisted document extraction to turn the contents into fields usable by the business process. Document classification can identify the document type so it reaches the appropriate workflow. Evaluate the intervention through reduced manual entry, processing time, and correction effort, while preserving required checks on the extracted information.

Trap Using a knowledge-search assistant to address an invoice-entry backlog, because finding answers does not itself populate the accounting workflow.

5 questions test this
Transaction risk detection creates value by directing attention to potentially fraudulent activity

When transaction patterns can reveal potentially fraudulent activity, use predictive fraud detection or anomaly detection to identify transactions needing proportionate investigation or additional checks. The business purpose is loss prevention, not generating explanations or content. Evaluate avoided fraud losses together with investigation effort and the friction imposed on legitimate customers; an unusual transaction is a reason to assess risk, not proof of fraud.

Trap Applying the same additional verification to every transaction, which adds customer effort without using transaction risk to target the checks.

4 questions test this
Build versus buy depends on differentiation, fit, and the ability to sustain delivery

When a mature commercial solution meets a common business need and time or specialist capacity is constrained, buying can accelerate value. Custom development is justified when distinctive requirements or competitive differentiation cannot be adequately met by available offerings and the organization can support the lifecycle. Compare functional fit, time to value, control, and ongoing capability rather than treating internal ownership as an advantage by itself.

Trap Building a commodity capability to preserve internal ownership, which consumes scarce delivery capacity without demonstrating a distinctive requirement.

5 questions test this
A partner can close a delivery capability gap while helping the organization learn

When an AI opportunity fits the strategy but internal expertise or delivery capacity is insufficient, a specialist partner can help co-build, integrate, and support the solution. Define deliverables, business acceptance criteria, knowledge transfer, and the eventual operating responsibilities. Partnership supplements capabilities; the organization still needs an accountable business owner and a plan for ongoing use.

Trap Treating a consulting engagement as the business ownership model, which leaves outcome accountability and the post-engagement handover undefined.

5 questions test this
Using an existing model does not prevent a business from creating a customized solution

A hybrid AI strategy can combine a provider's pretrained capabilities with company data, workflows, and a tailored application. This can preserve business-specific differentiation without the investment and specialist capability required to create a foundation model from scratch. Evaluate whether existing capabilities can satisfy the use case before funding foundational research and development.

Trap Treating a company-specific application as a requirement to create a new foundation model, which confuses application differentiation with foundational model development.

5 questions test this
A ready-to-use AI application and a managed AI development service solve different delivery needs

When a business needs ready-to-use workplace assistance, research, analytics, and workflow capabilities, evaluate an application such as Amazon Quick or suitable embedded AI in existing software. Amazon Bedrock provides managed foundation-model capabilities for teams creating tailored generative AI applications, while Amazon SageMaker AI supports teams developing and managing AI models. A managed development service reduces some technical burden but does not supply a finished business workflow by itself.

Trap Purchasing access to Amazon SageMaker AI as an immediately usable employee assistant, which provides model-development capabilities rather than the requested finished workplace application.

4 questions test this
Sourcing proposals must be compared over the same business scope and lifecycle

AI sourcing comparisons should include setup, integration, data preparation, licenses or usage, internal labor, training, human review, and ongoing support over a common evaluation period. A platform switch also adds transition work, temporary overlap, and any exit charges, which can delay or outweigh operating savings. Compare delivery commitments against the same must-have requirements and include all work needed to realize the proposed benefits; a low recurring price is not a whole-lifecycle business case.

Trap Comparing a vendor's subscription fee with a fully staffed internal implementation budget, which compares different cost boundaries and responsibilities.

5 questions test this
A proposed AI solution must be validated against the actual business workflow

When evaluating an AI purchase or platform replacement, validate representative tasks, permitted data, intended users, and agreed success criteria. Assess quality, usability, response time, volume requirements, integration effort, and operating cost in the complete workflow. During a platform switch, reassess existing prompts and integrations because equivalent task descriptions do not guarantee equivalent behavior. A controlled trial establishes business fitness more directly than a curated demonstration or a provider's general benchmark.

Trap Using a vendor's curated demonstration as the acceptance test, which does not establish fitness for the buyer's data and workflow.

5 questions test this
AI procurement must establish what the provider may do with business inputs and outputs

AI vendor due diligence must examine how prompts, source data, and outputs are stored, accessed, retained, and reused, including whether they may be used for model training. Establish the permitted data classifications and review ownership, commercial-use rights, and liability with legal stakeholders. Continue reviewing material contractual changes because a previously acceptable service can change its data-use or licensing terms.

Trap Using a paid consumer subscription as evidence of enterprise data protections, which does not establish the negotiated data-use and ownership terms the business requires.

5 questions test this
Vendor compliance fit depends on actual data flows and continuing evidence

When a use case has jurisdictional or regulatory constraints, verify where the vendor and its subprocessors store and process data and who can access it. Review relevant assurance evidence and contractual commitments against the organization's specific obligations. Procurement and compliance teams should reassess material changes rather than treating onboarding approval as a permanent guarantee.

Trap Using a vendor's general certification badge as proof of permitted cross-border processing, which does not establish the actual data locations and access arrangements.

6 questions test this
An AI exit strategy must account for portability, dependencies, and the cost of leaving

When evaluating a platform, establish whether business data and prompts can be exported in usable formats and identify provider-specific dependencies that would need replacement. Review exit procedures, assistance, associated charges, and the obligations that continue after termination. These findings inform both the initial sourcing choice and the feasibility of a later transition; a lower operating price alone does not establish a cheaper switch.

Trap Using a cancellation clause as evidence that a platform is portable, which addresses contract termination but not usable exports or replacement dependencies.

4 questions test this
AWS Marketplace simplifies finding and procuring AI offerings without replacing business evaluation

AWS Marketplace provides a catalog of software, data, and services with options for trials, product comparison, procurement, and consolidated AWS billing. Private offers can provide negotiated pricing and terms when public terms do not fit the requirement. The buyer must still assess suitability, data handling, and compliance; streamlined purchasing is not a substitute for due diligence.

Trap Treating a standard Marketplace listing as a negotiated agreement for bespoke commercial terms, which confuses a public offer with a seller's private offer.

5 questions test this
High potential value is insufficient when an AI initiative is not feasible or strategically relevant

Use a common prioritization matrix to compare business impact, strategic alignment, feasibility, risk, and delivery requirements. A high-value opportunity with usable data and manageable integration is a strong early candidate; one blocked by missing prerequisites needs enabling work before expansion. An easy technical demonstration should not outrank a better business opportunity solely because it is easier to show.

Trap Ranking projects by expected revenue alone, which ignores whether the organization can deliver the proposed outcome.

5 questions test this
AI portfolio sequencing must fund capabilities that later initiatives depend on

Some AI initiatives depend on earlier capabilities, such as reliable customer information or transaction categorization before more advanced customer recommendations. Sequence the enabling work before dependent outcomes and recognize its contribution to the wider portfolio. Evaluate dependencies across projects so an attractive downstream opportunity does not receive expansion funding while its essential foundation remains unavailable.

Trap Funding only the highest standalone ROI project, which can leave the prerequisites for realizing that project's benefits unfunded.

4 questions test this
A balanced AI portfolio combines near-term results with longer-term strategic options

AI portfolio management should balance small, demonstrable improvements with initiatives that may create new products or business models over a longer horizon. Quick wins build experience and confidence, while more uncertain opportunities need bounded learning investments and appropriate expectations. Compare initiatives within their intended horizons instead of requiring every strategic experiment to resemble an immediate cost-saving project.

Trap Applying an immediate operating-cost payback requirement to every exploratory investment, which can exclude strategically valuable options before their hypotheses are tested.

4 questions test this
Scale, pause, and termination decisions should follow explicit evidence-based gates

Set business success thresholds, quality limits, cost expectations, and readiness criteria before a pilot or release. Scale when the evidence meets the relevant gates, pause for a remediable gap with a defined next test, and stop or redirect when the use case cannot justify further investment. A convincing prototype is not sufficient evidence of business viability or readiness for full production.

Trap Using an executive deadline as the release criterion, which supplies schedule pressure rather than evidence that the agreed value and readiness gates were met.

4 questions test this
Sustainability is an AI investment consideration alongside business value and feasibility

AI proposals should account for the environmental resources required to deliver the intended benefit and the organization's sustainability goals. Compare alternatives that meet the business requirement and avoid funding unnecessary processing or capacity simply to maximize AI use. Track relevant efficiency or environmental measures alongside the outcome so a business improvement can be assessed against its resource footprint.

Trap Using financial ROI alone to rank proposals against an explicit environmental objective, which leaves the required sustainability outcome unmeasured.

4 questions test this
An AI portfolio must be reprioritized as business needs and observed results change

Review the AI portfolio regularly against current strategy, demonstrated outcomes, and resource constraints. Redirect future funding when an initiative does not address a priority or another opportunity provides stronger value for the available capacity. Prior investment explains history, but continued investment needs a current business rationale and realistic remaining benefits.

Trap Using the amount already spent as the reason to continue funding, which substitutes historical expenditure for the value of the next investment.

4 questions test this
Stable rules and well-defined paths usually favor deterministic automation

When a process has explicit rules, predictable inputs, and defined paths, use conventional automation if it satisfies the requirement. AI becomes useful when the work needs learned patterns, interpretation of varied information, or adaptation to context. Compare the flexibility actually needed with the additional cost, latency, and oversight of an AI approach.

Trap Using an autonomous agent for a fixed approval routing table, which adds reasoning and governance overhead to a process with fully specified transitions.

4 questions test this
Finding existing information does not automatically require generative AI

When users only need to locate an existing item or retrieve an exact approved value, a suitable search or lookup function may meet the need with fewer resources. Generative AI is more appropriate when the requirement includes synthesizing, explaining, or drafting content from information. Evaluate the required outcome rather than replacing a working retrieval task merely because generation is available.

Trap Using generated prose as an exact account-balance lookup, which adds probabilistic generation to a requirement for an authoritative stored value.

4 questions test this
An AI approach is not ready when its essential data is unusable or inaccessible

An AI proposal needs data or knowledge sources appropriate to the intended task, with adequate quality and permitted access. When these prerequisites cannot be met, first remediate the constraint, narrow the use case, or select a workable non-AI process. A sophisticated solution cannot establish missing business facts or make prohibited data use acceptable.

Trap Collecting a larger volume of the same unreliable records as the remedy, which increases quantity without fixing the essential quality or access gap.

4 questions test this
AI is unsuitable for an autonomous decision when its errors cannot be acceptably controlled

AI results are probabilistic, so suitability depends on the consequence of error and the available controls. For consequential work, consider a human-reviewed assistive role if it can deliver value within the required oversight. If the process requires absolute predictability or the necessary review and safeguards are not feasible, use an appropriate deterministic or human process instead of forcing autonomous AI into the role.

Trap Using strong average accuracy as authorization for unreviewed consequential decisions, which does not establish that the remaining errors are acceptable or controlled.

4 questions test this
A staged AI transition limits business exposure while the replacement proves itself

When replacing an established process or platform, introduce the new capability to a limited population or workflow before expanding. Define service acceptance criteria, ownership of exceptions, and a usable fallback or return path for unacceptable results. Retain the current capability until the business has evidence that the replacement can support the required work without unacceptable disruption.

Trap Retiring the current process immediately after a successful demonstration, which removes continuity protection before representative operational validation.

4 questions test this
A platform transition must preserve usable information and appropriate access

Before moving an AI-enabled process, inventory the information sources, formats, refresh needs, ownership, and access rules that make it work. Confirm that the destination can use the required information and preserve the relevant restrictions and freshness. Resolve these readiness gaps before business acceptance because moving the application does not itself establish that its knowledge and permissions remain correct.

Trap Approving the transition after copying the documents alone, which does not establish that the destination can use current content under the required access rules.

4 questions test this
Moving a process to AI changes human work as well as software

An AI process transition must define who reviews outputs, handles exceptions, approves consequential actions, and supports users. Prepare affected employees for the new workflow before shifting the work, and include that effort in readiness and cost decisions. Buying or switching technology does not automatically establish a workable division of responsibility between the AI capability and people.

Trap Using tool access as the employee transition plan, which supplies software availability without establishing new responsibilities or exception handling.

4 questions test this
AI continuity depends on every essential part of the business service

AI service continuity requires more than an available model: the process also depends on its information sources, connected business systems, support arrangements, and recovery capability. Identify essential dependencies and obtain evidence that the recovery arrangements can support the required business service. A provider availability commitment does not by itself demonstrate that the complete workflow can recover after a disruption.

Trap Using a model provider's uptime commitment as the complete continuity plan, which leaves the workflow's data and connected-system dependencies unassessed.

4 questions test this

Measure and demonstrate AI business value

Read full chapter
  • AI KPIs must connect system performance to a business outcome
  • Efficiency gains need quality and customer-experience guardrails
  • Nonfinancial AI benefits still need observable evidence
  • Acceptable AI performance depends on which errors matter to the business
  • A KPI needs a consistent definition and an owner who can act on it
  • Defensible AI targets combine current performance with relevant benchmarks and user acceptance
  • The baseline must describe the process AI is intended to improve before the change
  • Volume-normalized measures distinguish efficiency from changes in workload
  • A comparable non-AI group helps separate AI effects from other business changes
  • Performance comparisons need a stable, representative evaluation basis
  • ROI measures net benefit relative to the investment cost
  • Payback measures how long it takes cumulative net benefits to recover the initial investment
  • A fair AI return comparison includes one-time and recurring amounts over the same horizon
  • The labor value of AI time savings depends on net effort, volume, and the relevant labor rate
  • Revenue uplift contributes value after the costs of delivering the extra sales
  • Lower per-task AI costs justify higher fixed costs only after sufficient work is completed
  • Benefits received later can be worth less today than equal benefits received sooner
  • Released staff capacity, avoided future cost, and reduced current spending are different benefits
  • An AI business case must not count the same economic gain more than once
  • AI return estimates should be tested under plausible changes in adoption, costs, and benefits
  • AI ROI is an ongoing management measure rather than a one-time launch calculation
  • Error reduction creates financial value through avoided rework and business losses
  • Adaptability can create value through avoided rebuilds and downtime without eliminating maintenance
  • Repeated use and task completion are more informative adoption signals than access alone
  • Acceptance, rework, and feedback patterns can reveal AI value problems early
  • Readiness evidence provides early warning when an AI project is unlikely to reach business use
  • Cost and quality trends can warn that an AI initiative will miss its value case
  • Consumption-based AI pricing depends on the amount and type of work performed
  • Instance-based AI costs depend on resource type and running time
  • Seat-based AI pricing does not necessarily make the total bill fixed or usage unlimited
  • AI purchase commitments are attractive only when expected use justifies the obligation
  • AWS Pricing Calculator estimates proposed scenarios while AWS Cost Explorer analyzes actual usage and spending
  • A pilot cost forecast must account for production volume and workload shape
  • An AI solution's bill extends beyond its model or application charges
  • AI cost accountability requires attribution to the work and business unit driving consumption
  • Showback creates cost visibility while chargeback assigns costs to consuming business units
  • Phased funding can support early AI experimentation and later usage accountability
  • AI financial guardrails require active review and a defined response to deviations
  • AI cost optimization seeks the lowest cost that still delivers the required business outcome
  • Controlling unnecessary AI work reduces cost without requiring less useful service
  • AI work that can wait may qualify for lower-cost batch processing
  • Prompt caching is economical when repeated context avoids enough processing to justify cache costs
  • Stopping an AI initiative does not stop charges until its unused resources and commitments are addressed
  • A proven AI use case should test whether a smaller model can meet the same business requirements
  • A continuing AI initiative can reduce waste by matching paid capacity to actual need

Unlock with Premium — includes all practice exams and the complete study guide.

Position AI for competitive advantage

Read full chapter
  • Competitive AI assessment should compare customer-relevant capabilities rather than technology labels
  • An industry AI success story is a hypothesis to validate, not a transferable guarantee
  • AI can improve competitive position through lower cost to serve or distinctive customer value
  • Market and operating signals help determine which AI opportunities are timely
  • AI competitive assessment must consider new ways to deliver value, not just current rivals
  • Business-model transformation changes how value is offered or captured
  • AI-enabled predictive service can support new recurring offerings around a product
  • Data monetization includes internal business improvement before external sale
  • AI can make new markets reachable by expanding the service capacity of an existing team
  • A strategic AI partnership can create routes to market as well as implementation capacity
  • Outcome-based pricing ties payment to an agreed successful business result
  • Widely available AI capabilities are not a durable competitive advantage by themselves
  • A data flywheel compounds value only when useful new data improves the service
  • Responsible AI can support differentiation through customer and employee trust
  • AI-enabled advantage must adapt as customer expectations and competitors change
  • Demand prediction creates operational value when it improves inventory and service decisions
  • Predictive maintenance creates value by enabling action before an operational failure
  • AI-assisted inspection improves operations when defect detection leads to better quality outcomes
  • Reusable AI capabilities reduce duplicated effort as adoption expands
  • AI investment intensity should reflect market opportunity and the company's ability to act
  • Organizational maturity changes which AI investments are most useful
  • An AI capability can merit investment to maintain competitive parity without being a durable differentiator
  • The cost of waiting must be weighed against the risk of investing too broadly
  • AI investment decisions must account for the opportunities displaced by scarce resources

Unlock with Premium — includes all practice exams and the complete study guide.

AI Governance and Responsible AI Leadership

Apply Responsible AI Principles to Business Decisions

Read full chapter

Cheat sheet

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

Fairness goals must reflect the people and outcomes a use case affects

Fairness considers impacts on different stakeholder groups, not just whether an AI system treats every input identically. Similar treatment of comparable individuals and appropriate outcomes across groups answer different fairness concerns; choose the goal with affected stakeholders and document conflicts between goals. Access also matters: balanced prediction rates do not make a service fair if people with disabilities cannot use it.

Trap Using identical interaction requirements for every customer as the fairness standard, which overlooks accessibility and differences in stakeholder needs.

5 questions test this
An explanation of one decision serves a different need from an explanation of overall behavior

Explainability helps people understand why an AI system produced an output. When an applicant needs reasons for one denial, require an explanation of that individual decision; when an oversight team needs to understand general behavior, require information about the factors influencing predictions across cases. Set the required explanation level with business and compliance stakeholders before committing to a solution.

Trap Providing a portfolio-wide factor summary to explain an individual denial, which does not establish what influenced that applicant's decision.

3 questions test this
Privacy governs appropriate data use while security protects data and systems

Privacy concerns whether information is appropriately obtained, used, and managed; security concerns protection against unauthorized access, manipulation, and disclosure. An authorized employee can still use personal information for an inappropriate purpose, so a privacy review is needed even when access is secure. Conversely, a permitted business purpose still requires protection against theft and tampering.

Trap Using encryption approval as the authorization for a new personal-data purpose, which evaluates protection rather than whether that use is permitted.

4 questions test this
Factual correctness does not establish that an AI output is safe

Safety addresses harmful outputs, actions, and misuse, whereas veracity addresses factual correctness. Accurate instructions can still enable harm, so a business must assess permitted uses as well as truthfulness. A harmless-sounding answer can also be factually wrong, requiring factual validation even when it passes a harmful-content check.

Trap Using a factual-accuracy review as the safety approval, which does not assess harmful uses of accurate information.

4 questions test this
Robustness means maintaining appropriate performance as inputs vary

Robustness concerns correct outputs for expected and unexpected inputs, including ordinary changes in language, image quality, or environmental conditions. Require evaluation on the variations actual users will encounter, not only clean examples. Distinguish changes that should affect the answer from irrelevant variations the system should tolerate, and assess combinations such as accented speech with background noise.

Trap Using higher service uptime as the remedy for dialect-dependent errors, which measures availability rather than performance under input variation.

4 questions test this
Controllability requires interventions that reliably change AI behavior

Controllability means having mechanisms to monitor and steer system behavior, including adjustment, override, or rollback where appropriate. Judge a control by whether it produces the intended change quickly enough for the business risk, and test whether it creates unintended effects elsewhere. A visible control is insufficient evidence unless users can operate it effectively and understand when it cannot work.

Trap Using a monitoring dashboard as the complete controllability mechanism, which provides visibility without an effective way to intervene.

4 questions test this
Equal approval rates and equal recognition of qualified people are different fairness goals

Demographic parity asks whether groups receive positive decisions at comparable rates, such as similar shares of all applicants being recommended for a job. Equal opportunity asks whether genuinely qualified applicants are correctly recommended at comparable rates across groups. Because the first comparison includes all applicants and the second includes qualified applicants, meeting one goal does not establish the other; agree which concern matters in the business context before judging the results.

Trap Using similar overall accuracy across groups to assess equal recognition of qualified candidates, because overall correctness also counts negative cases and can conceal differences in identifying qualified people.

AI disclosure should let people make informed choices before relying on the system

Transparency enables people to make informed choices about engaging with AI. Identify where users first encounter the system and where they act on its outputs, then disclose AI involvement and relevant limitations at those points. Match the prominence and frequency of disclosures to the context and risk rather than relying only on a distant policy page.

Trap Placing the only AI notice in an annual corporate report, which does not inform users at the points where they engage with or rely on AI.

4 questions test this
System cards communicate intended uses and limitations but do not replace local evaluation

An AI system card brings together intended uses, limitations, responsible AI design choices, and relevant performance information. Use it to judge whether a proposed application fits the documented boundaries and to prepare guidance for downstream users. The deploying organization must still evaluate performance on its own inputs and workflow; supplier documentation does not establish suitability for every customer context.

Trap Using a supplier's general service card as the go-live evidence for a new population, which omits evaluation of the deployer's specific inputs and conditions.

4 questions test this
Useful transparency is tailored to the audience and protects sensitive information

A transparency strategy should match each stakeholder's information needs, expertise, accessibility requirements, and communication channel. Customers may need plain-language capabilities and limitations, while authorized reviewers need more detailed evidence. Review public disclosures separately from restricted materials so that openness does not reveal personal data, proprietary information, or exploitable security details.

Trap Publishing raw internal evaluation records to maximize transparency, which can expose protected information without giving users an understandable explanation.

4 questions test this
AI user guidance must change when capabilities, limitations, or intended uses change

AI documentation needs an owner, review process, version history, and a way to reach affected users. Trigger review after system releases, new uses, significant stakeholder feedback, or relevant changes in expectations. Updating an internal document is only part of the task: publish the approved guidance through the channels users actually rely on and address outdated copies.

Trap Updating the engineering release notes alone, which does not revise or distribute the guidance business users need for responsible use.

4 questions test this
Uncertainty information is useful only if users know when to rely on or question an output

AI guidance should explain uncertainty, known limitations, and the circumstances that call for additional verification. Evaluate whether different user groups actually recognize when an output deserves trust, further checking, or rejection. Measure comprehension and decision quality as well as reported trust, because a persuasive interface can increase confidence without improving judgment.

Trap Using higher user trust ratings as the sole proof of effective explanations, which does not show that users can identify unreliable recommendations.

4 questions test this
Business performance improvements cannot substitute for mandatory obligations

When an AI use case has a mandatory explanation, privacy, or other compliance requirement, treat that requirement as a condition of proceeding. Accept a lower business-performance result if it is the feasible compliant option and still meets necessary operating standards. If no option meets the obligation, change the scope or defer the use rather than promising compliance after release.

Trap Selecting the highest-accuracy option with a later compliance retrofit, which leaves the mandatory obligation unmet when the system begins making decisions.

4 questions test this
A responsible AI tradeoff considers who receives the benefit and who bears the harm

Evaluate benefits and harms separately for affected stakeholder groups rather than treating company-wide gains as a complete justification. Include people indirectly affected by decisions and groups that may experience disproportionately severe harm. Use explicit organizational principles and stakeholder needs to prioritize competing outcomes, and document the rationale so later decisions use the same standards.

Trap Using total company savings as the complete tradeoff criterion, which leaves the distribution and severity of stakeholder harms unexamined.

4 questions test this
Privacy-enhancing choices must be evaluated alongside utility and fairness

Privacy protections such as minimizing, de-identifying, or aggregating information can reduce exposure but may also reduce useful detail. Evaluate the effect on the business task and on different groups, particularly where data is sparse. Choose the least sensitive information that can support the authorized purpose, and document acceptable utility tradeoffs without relaxing mandatory privacy obligations.

Trap Maximizing data aggregation as the only design goal, which can remove detail needed for useful or fair outcomes in smaller groups.

4 questions test this
Safety controls should be assessed for missed harms and unnecessary blocking

Protective filters can miss harmful material and can also block legitimate requests. Assess both kinds of error against the use case, user population, and harm severity before agreeing acceptance criteria. A stricter control may be justified for severe harms, while ambiguous legitimate requests may need review rather than automatic rejection; monitor useful service delivery alongside protection effectiveness.

Trap Using the percentage of blocked requests as the sole safety success metric, which rewards unnecessary blocking as well as genuine protection.

4 questions test this
Narrowing an AI use case can preserve value when broad deployment is not supportable

When a system cannot meet essential criteria across a broad application, restrict it to domains, inputs, locations, or users for which it can meet the required standards. Reassess benefits and risks and update the criteria for that narrower use. Scope reduction is a valid risk treatment when meaningful value remains; persistent inability to control harm is a reason to stop that use, not lower essential standards.

Trap Lowering the safety acceptance threshold to preserve the original rollout scope, which changes the protection standard instead of resolving the unsupported use.

4 questions test this
Responsible AI belongs in initial planning and continues throughout the lifecycle

Governance by design incorporates responsible AI requirements when defining the use case, before data and solution commitments lock in avoidable risks. Carry those requirements into evaluation, user guidance, production monitoring, and retirement. Iterative pilots can still move quickly, but their boundaries and required approvals must be clear from the start rather than postponed until production.

Trap Scheduling responsible AI review only after the pilot is selected for production, which allows consequential data and workflow decisions to precede the relevant requirements.

5 questions test this
AI planning must include upstream contributors and people indirectly affected by outputs

Stakeholder mapping should cover direct users, people whose information is processed, people affected without using the system, and upstream contributors such as suppliers. Seek perspectives outside the project team, including vulnerable populations where relevant, to reveal missed requirements and harm scenarios. Revisit the map when the workflow or audience changes rather than treating the initial list as permanent.

Trap Surveying only paying users to establish stakeholder requirements, which omits data subjects, indirect stakeholders, and upstream dependencies.

5 questions test this
Responsible AI requirements need measurable pass-or-fail release criteria

Translate expected benefits and identified harms into clear release questions, measurements, and acceptance thresholds before building or buying around the use case. Use existing performance evidence, stakeholder requirements, and risk tolerance to set realistic targets, and identify who is affected by each failed criterion. A complex promise such as a good user experience may need separate criteria for correctness, response time, and usability; no single universal accuracy percentage establishes responsible readiness.

5 questions test this
Risk mitigation can combine safer system behavior, protective filtering, and user guidance

Assign each identified harm to a mitigation strategy: improve the underlying behavior, screen inputs or outputs, guide users, or combine these approaches. Match the strategy to how the harm arises and how much users can reasonably control. Guidance is valuable for appropriate use, while risks requiring enforced boundaries need system safeguards as well; evaluate the resulting protection rather than assuming one strategy fits every harm.

Trap Using a user disclaimer as the sole control for a prohibited system action, which asks users to respect a boundary that the system still permits.

5 questions test this
Critical or subjective AI evaluations benefit from an independent second assessment

When safety, fairness, or another consequential judgment depends heavily on interpretation, obtain corroboration from qualified reviewers outside the delivery team's incentives. For higher-risk uses, an independent team can develop evaluation cases from the documented intended use rather than relying entirely on cases chosen by the supplier or project team. Investigate material disagreements before approval and keep both findings as decision evidence.

Trap Repeating the delivery team's own demonstration as an independent review, which retains the same case-selection assumptions and incentives.

5 questions test this
Evaluation evidence must identify the system and data versions actually being approved

System updates require renewed evaluation and traceable records of the version, evaluation data, conditions, methods, and results. Recheck that criteria still cover current capabilities and relevant expectations instead of carrying forward approval for a materially different system. Link evaluation evidence to the release decision so an auditor can reconstruct what was tested and why that version was accepted.

Trap Using the previous version's safety report for an expanded release, which provides evidence about a different system and possibly different risks.

5 questions test this
Automated evaluation scales assessment but its judgments still need validation

Choose automated evaluation for repeatable measurements it can capture reliably, and use qualified human assessment for context or experience the automated method misses. An AI system used to judge another AI system should be checked against human judgments before its scores become approval evidence. Human assessment also needs clear criteria and consistency checks; neither automation nor expert involvement makes an evaluation method automatically reliable.

Trap Using an AI evaluator's high scores as self-validating approval evidence, which omits whether its judgments match the intended quality standard.

5 questions test this
Human oversight should be placed where expertise can prevent consequential errors

Map the business workflow to identify decisions whose errors are hard to judge or have serious consequences. Require human review before an effect occurs when later correction would be inadequate; lower-risk, reversible activity may support automated handling with supervisory review. Oversight placement depends on the actual consequence and the value of human expertise, not merely whether a system is customer-facing.

Trap Using a monthly retrospective sample as the only oversight for irreversible individual decisions, which places review after the point at which it could prevent harm.

5 questions test this
Human review is meaningful only when reviewers can assess and overturn AI recommendations

Human reviewers need relevant expertise, training, sufficient information, and authority to reject or correct an AI recommendation. Give them the input evidence, explanations, and limitations needed for independent judgment, not just an approve button. Evaluate the human-and-AI workflow as well as the system itself so that nominal review does not become routine acceptance of unreliable outputs.

Trap Adding mandatory approval clicks for staff who cannot inspect the evidence or overturn the recommendation, which creates nominal rather than effective oversight.

4 questions test this
Escalation rules must specify both the trigger and the person authorized to act

Define escalation criteria using factors such as harm severity, uncertainty, policy violations, and business impact. Specify the reviewer, required response, decision authority, and documentation so that an exception reaches someone able to resolve it. High-impact policy violations may require escalation even when an output appears confident, while ambiguous lower-risk cases may follow a different review path.

Trap Routing every exception to a general inbox without decision rights, which supplies notification but no accountable resolution path.

4 questions test this
People affected by AI decisions need a way to challenge and correct outcomes

A recourse process lets affected people report problems, challenge an outcome, and obtain review of incorrect information or decisions. Make the route understandable and accessible and give reviewers the ability to provide an appropriate remedy. Feed validated complaints into system evaluation as well, because resolving one case and identifying a recurring problem serve different needs.

Trap Offering only a satisfaction survey after an adverse decision, which gathers opinions without providing review or correction of that decision.

4 questions test this
Agent autonomy requires explicit boundaries on actions taken for the organization

When AI can act rather than merely recommend, define which activities it may perform autonomously and which require human approval. Set approval boundaries before irreversible or consequential effects and extend oversight records to actions and external interactions, not just chat responses. An approved information assistant does not automatically have approval to make commitments or change business records on the company's behalf.

Trap Extending a chatbot's approval to autonomous business transactions, which treats permission to provide information as permission to act.

4 questions test this
Guardrails turn AI policy boundaries into enforced checks

Guardrails are safeguards that enforce permitted boundaries on AI inputs, outputs, or actions, such as blocking prohibited content or preventing an action until a reviewer approves it. Recognize them as the enforcement layer that supports the organization's mitigation policy, rather than as another name for user guidance or human judgment. Guardrails complement human oversight: unfamiliar or consequential cases still need the appropriate review and escalation path.

Trap Using a system prompt as the sole policy-enforcement control, because instructions to the model do not independently verify compliance.

4 questions test this

Establish AI Governance and Regulatory Compliance

Read full chapter
  • AI governance needs cross-functional representation and explicit decision authority
  • Shared AI work still needs named owners and clear handoffs
  • A centralized AI inventory should make ownership and permitted use visible
  • Managing shadow AI requires a usable approved path as well as policy enforcement
  • AI governance should extend existing enterprise controls rather than operate in isolation
  • Buying AI changes the delivery arrangement but does not remove the need for supplier governance
  • AI compliance depends on where the system operates and how the organization uses it
  • A privacy impact assessment examines the proposed processing before risks become embedded
  • Permission to hold customer information is not blanket permission to reuse it for AI
  • AI-generated information about an identifiable person can carry privacy obligations even when it is wrong
  • AI vendor review must follow data beyond the company's immediate interface
  • Compliance evidence should connect obligations, assessments, controls, and approval decisions
  • Legal AI risk categories impose duties that an internal risk score cannot remove
  • AI access should be limited to the data and functions each role needs
  • Strong identity controls and appropriate permissions solve different access problems
  • Encryption protects stored and transmitted AI data but must be combined with access controls
  • Sensitive AI information should be minimized before use and screened for inappropriate disclosure
  • Monitoring records need their own access, retention, and privacy controls
  • AWS infrastructure protection does not replace the customer's AI data responsibilities
  • Prompt injection can turn untrusted material into attempted instructions
  • Data poisoning targets information the AI system learns from or relies on
  • Information sensitivity determines the handling rules for AI use
  • Risk classification combines likelihood with the severity and reversibility of harm
  • The business use and degree of autonomy shape an AI system's risk
  • Risk remaining after controls must be assessed against the organization's tolerance
  • Governance effort should be proportional to risk and able to stop unacceptable uses
  • An AI risk register should track harms and treatment evidence across use cases
  • AI risk ratings require review as the system and its operating context evolve
  • ISO/IEC 42001 provides requirements for an organizational AI management system
  • ISO/IEC 23053 provides a common framework for describing AI systems that use machine learning
  • The NIST AI RMF connects Govern, Map, Measure, and Manage rather than treating risk as a one-time checklist
  • AWS CAF for AI places responsible use within a broader organizational adoption framework
  • The NIST Generative AI Profile applies the AI RMF to risks amplified by generative AI

Unlock with Premium — includes all practice exams and the complete study guide.

Identify Enterprise AI Risks and Direct Mitigation

Read full chapter
  • Production AI monitoring should cover business outcomes and responsible AI risks
  • An operating baseline describes normal behavior while acceptance limits define what is allowed
  • Automated AI monitoring needs periodic checks of its own effectiveness
  • AI incident plans must connect containment, accountable response, and learning
  • Monitoring creates value only when findings lead to owned and validated improvements
  • AI retirement must address dependent workflows and continuing data obligations
  • Changes in what drives AI decisions deserve production review
  • AI failure reviews need evidence tied to the interaction that failed
  • Bias can arise from data, system development, evaluation, and human use
  • Representative data must cover the groups and situations the AI use will encounter
  • Past decisions and automated labels are not automatically fair ground truth
  • Removing explicit demographic fields does not establish that an AI system is unbiased
  • Overall AI performance can conceal poor outcomes for particular groups
  • Fairness can deteriorate after launch even when the AI system itself is unchanged
  • Human participation can introduce bias as well as provide a safeguard
  • Bias drift names a change in fairness after deployment
  • AI decisions can reshape future evidence and reinforce an existing disparity
  • Bias remediation should target the lifecycle stage producing the disparity
  • Harmful-content policy must reflect the use case, audience, and forms of output
  • Input screening and output screening protect different parts of an AI interaction
  • Topic restrictions, word matching, and sensitive-information controls address different problems
  • Safety evaluation needs realistic ordinary use and deliberate attempts to expose weaknesses
  • Human content review requires protections for reviewers as well as users
  • Data access and data-use rights must be assessed separately for AI
  • Generated content can raise third-party rights concerns even when no copying was requested
  • AI supplier contracts should define content rights and operational responsibilities explicitly
  • Content provenance helps trace origin and changes but does not prove a claim is true
  • AI intellectual-property review must include the company's own confidential information
  • AI content governance needs a process for investigating and responding to rights claims
  • Synthetic-content markers disclose AI origin rather than certify the message
  • AI factual errors include fabricated content, omissions, and misleading emphasis
  • An answer can be grounded in a source yet fail to answer the user's question
  • Reliable prediction monitoring uses observed outcomes when they become available
  • Grounding improves reliability only when the supporting information is trustworthy and current
  • Data-quality degradation and data drift need different first responses
  • Small AI errors can become larger risks when downstream steps rely on them
  • AI continuity plans need tested fallback processes, including manual handling where appropriate
  • Evaluation overlap can make AI reliability appear stronger than it is
  • Hallucination detection checks the support behind an AI assertion
  • Model drift can reflect changed business relationships even when inputs look familiar

Unlock with Premium — includes all practice exams and the complete study guide.

Business Readiness, Leadership, and AI Transformation

Assess AI business readiness and maturity

Read full chapter

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

Establish data and infrastructure foundations for AI

Read full chapter

Cheat sheet

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

Data quality dimensions identify different kinds of readiness defects

Data accuracy asks whether values are correct, completeness asks whether required information is present, and consistency asks whether representations agree across records and systems. Timeliness asks whether data is current enough for its intended use. Evaluate the dimensions that matter to the business process so a dataset's strength in one dimension does not conceal a different defect.

Trap Using a high record-count target as the data acceptance criterion because volume does not establish correctness or completeness.

4 questions test this
Reliable AI inputs require consistent collection and source-level quality controls

When sites collect the same business information differently, standardize collection practices, definitions, and quality checks before expanding an AI initiative. Assign responsibility for recurring defects and improve the collection process at their source. Consistent controls across locations make reliable inputs an ongoing capability instead of a one-time cleanup.

Trap Using centralized output correction to compensate for inconsistent store inputs because it leaves the recurring source defect in place.

4 questions test this
Data must represent the people and situations the AI initiative will serve

Data readiness includes relevance to the use case and representation of the intended users, operating conditions, and business situations. A large, internally consistent dataset can still omit important populations or contexts. Evaluate coverage of the target setting and potential representational bias before treating existing records as an adequate foundation.

Trap Using clean historical records as the default representative dataset because cleanliness does not establish relevance to the target population.

4 questions test this
Required data freshness depends on how quickly the business information changes

Set data update expectations according to the use case's need for current information. A dynamic knowledge source or time-sensitive decision process needs a different refresh approach from a stable reference collection. Include update frequency and quality monitoring in readiness requirements so data that was suitable at the pilot's start remains usable later.

Trap Using the document archive's publication schedule for a live policy assistant because that schedule may not meet its currency requirement.

4 questions test this
Uniqueness and validity reveal defects that correct-looking values can conceal

Uniqueness checks for unwanted duplicate records of the same business entity or event; duplicates can inflate counts and give repeated evidence undue weight even when their individual values are correct. Validity asks whether values meet the required formats and business rules, such as an accepted identifier format or a permitted category. Include duplicate rate and the share of values passing validity rules in the data-readiness scorecard, with acceptance criteria tied to the intended business use. Passing a format rule does not establish that a value is factually accurate, and an accurate value does not establish that its record is unique.

Trap Using a high required-field completion rate as sufficient evidence of readiness when duplicate customers and invalid identifiers remain.

4 questions test this
Fragmented enterprise data requires a unified access and governance strategy

When AI must use information across disconnected systems, establish a unified data strategy that connects and governs the required sources. Unified access can work across existing locations; it does not require every asset to reside in one physical repository. Resolve fragmentation in access, formats, and governance as an enterprise foundation rather than leaving each initiative to negotiate an isolated solution.

Trap Using additional processing capacity to expand a single-database pilot because capacity does not connect disconnected information sources.

5 questions test this
Data democratization means easier authorized access, not unrestricted access

A data sharing framework should make relevant data discoverable and accessible to appropriate users while preserving privacy and use restrictions. Define clear access-request workflows and role-appropriate permissions so routine legitimate use does not depend on ad hoc negotiation. Excessive control can recreate silos, while excessive access exposes information beyond its permitted audience.

Trap Using an organization-wide read permission to accelerate collaboration because broad availability does not establish authorization for every user.

4 questions test this
Connecting datasets is insufficient when their business definitions conflict

Shared business glossaries, data dictionaries, and reference definitions give teams a consistent interpretation of enterprise data. Reconcile meanings such as customer status, product identifiers, and reporting periods before using information together. Technical connectivity solves access, whereas shared semantics make the combined information interpretable and comparable.

Trap Using a common file format as proof of interoperability because matching formats can still carry conflicting business meanings.

5 questions test this
An AI data strategy must account for useful unstructured information

Enterprise AI can need documents, messages, images, or other unstructured material as well as structured business records. Include the relevant formats in data discovery, classification, quality assessment, and authorized access planning. Plan how users and AI applications will find and understand this information's context while preserving appropriate controls.

Trap Using the existing reporting warehouse inventory as the complete data assessment because important business knowledge may reside outside tabular systems.

5 questions test this
Data silos restrict usable business context and create reconciliation work

Data silos are isolated systems or departmental collections whose information is difficult for other authorized users or applications to access or share. They limit the business context available to AI and can create duplicate or inconsistent records that require manual transfer and reconciliation. Diagnose whether the isolation comes from disconnected technology or from organizational barriers such as ownership disputes and unwillingness to share, because technical connectivity alone does not resolve the latter.

Trap Funding model retraining to compensate for business records that remain inaccessible to the AI application.

5 questions test this
Data strategy should work backward from business initiatives

A data strategy should identify which domains, sources, quality levels, and access capabilities are needed for targeted business initiatives. Translate the business requirements into a data roadmap and measures of whether the required information is fit for its intended purpose. This makes the data foundation an enabler of business value rather than a collection program seeking a future purpose.

Trap Using total enterprise data collected as the data strategy's success measure because collection volume does not establish support for business priorities.

4 questions test this
Data ownership and stewardship provide different forms of accountability

Data owners define policies for their data, including access conditions and key business definitions, while data stewards apply governance in day-to-day work, identify issues, and escalate them. Technical teams provide the supporting mechanisms, but business responsibilities must still be explicit. Route policy decisions to the responsible owner and recurring quality or governance issues through the stewardship process.

Trap Using the storage administrator as the default business data owner because technical custody does not establish authority over permitted business use.

4 questions test this
Federated data governance combines domain responsibility with enterprise consistency

When business units need local data expertise and autonomy, federated governance lets domains manage their data while a central function maintains shared policies and common capabilities. Preserve coherence across domains so information can still be linked and used responsibly. Centralized governance may fit common control needs, but local ownership need not mean independent, incompatible standards.

Trap Using separate departmental dictionaries and access rules as the definition of federation because incompatible rules undermine cross-domain use.

4 questions test this
A data sharing framework must define the conditions of use

Data sharing rules should specify who may use which data, for what purposes, and under what protection and compliance conditions. Assess privacy, confidentiality, consent, and applicable use restrictions alongside availability and quality; sensitive information may require protection or exclusion. Make ownership, approval, and ongoing responsibilities clear across organizational boundaries so authorized access is paired with an authorized AI purpose.

Trap Using a successful interdepartmental file transfer as evidence of an adequate sharing agreement because transport does not define permitted reuse.

4 questions test this
A data catalog helps users find and understand candidate data

A data catalog inventories data assets and their metadata, including ownership, meaning, provenance, and quality information. It supports discovery and informed access requests instead of forcing each AI team to rediscover sources. Treat the catalog as part of governance: a listing alone does not supply the policies, stewardship, and quality processes that make data trustworthy.

Trap Using folder names as the enterprise data catalog because locations alone do not explain ownership, meaning, or fitness for use.

4 questions test this
Lineage and version records make AI data use traceable

Data lineage records where information originated and how it changed, while version records identify the particular dataset state used. Together they support impact assessment, investigation of quality problems, and reproduction of prior results. A dataset can be suitable for operational use without being sufficient evidence of what supported an earlier AI output.

Trap Using the source department's name as the lineage record because ownership metadata does not describe transformations or the dataset state used.

4 questions test this
AI data needs lifecycle rules rather than indefinite accumulation

Data governance should define lifecycle policies for retaining, updating, and disposing of data in line with business needs and compliance obligations. Include AI-related data assets in those policies so old copies do not become unmanaged exceptions. Retention supports legitimate operational and evidence needs, while disposal controls data without a justified use.

5 questions test this
Infrastructure requirements must reflect workload demand and service expectations

Assess whether the foundation can handle the expected data volume, arrival rate, concurrent demand, and required response speed. A real-time interaction needs different responsiveness from a scheduled batch task, and a successful low-volume demonstration does not establish peak-load capacity. Ask for a scalable capacity approach that balances performance with cost rather than assuming the largest possible environment is necessary.

Trap Using average daily request volume alone to approve capacity because it can conceal short periods of high concurrent demand.

5 questions test this
An AI capability needs a workable connection to the business process

Assess how the AI solution will obtain necessary information and deliver results into the systems and workflows people already use. Compatibility, authorized data access, and operational handoffs are foundational requirements, not optional additions after a demonstration. A standalone tool can prove a capability without proving that the organization can use it effectively in its end-to-end process.

Trap Using a standalone application's demonstration score as integration acceptance because it does not establish usable connections to the business workflow.

5 questions test this
Managed AI platforms can reduce the foundation an enterprise must build itself

When assessing foundational needs, distinguish access to managed foundation models from a broader requirement for custom machine learning capabilities. Amazon Bedrock provides managed access to foundation models for generative AI applications, while Amazon SageMaker AI supports building, training, and deploying machine learning models. Match the platform category and supporting skills to the initiative instead of assuming every AI project needs an internally built model environment.

Trap Using a custom model-development program for every generative AI use case because some initiatives need managed model access rather than new model construction.

5 questions test this
A usable AI foundation needs protection and observability as well as capacity

Foundational assessment should include the ability to protect sensitive data, govern access, record relevant activity, and observe service and AI quality. Computational capacity cannot establish these capabilities, and a managed service does not remove the need to assess them for the complete application. State the business requirements and obtain evidence from the responsible technical and risk teams without treating infrastructure configuration as the business strategist's task.

Trap Using a compute performance benchmark as the complete platform acceptance test because it says little about protection or operational visibility.

5 questions test this

Lead AI change and build workforce capabilities

Read full chapter
  • Executive sponsorship requires decisions and visible support
  • Business and technology leaders should co-lead AI transformation
  • AI champions turn organizational intent into local adoption
  • Leadership alignment must be maintained as the initiative evolves
  • AI teams need business, technical, and risk expertise from the start
  • A RACI matrix makes participation and decision rights explicit
  • Business outcome accountability should remain explicit alongside technical ownership
  • Workforce and supplier decisions require their own functional expertise
  • AI change communication should explain why the change matters and what it changes
  • Different stakeholder groups need different AI change messages and channels
  • Two-way communication lets employees understand and influence the change
  • A communication strategy sets direction while the plan schedules delivery
  • Psychological safety supports learning without removing performance expectations
  • Rewards and management practices should reinforce the desired AI culture
  • Change overload calls for sequencing and capacity adjustments
  • Risk aversion becomes an adoption barrier when uncertainty rules out bounded learning
  • Resistance to change needs an intervention matched to the underlying concern
  • AI literacy gives business and technical teams a shared language
  • Workforce development should target the skills each role needs
  • Hackathons and proofs of concept serve different workforce learning needs
  • Responsible AI training is part of workforce readiness
  • AI proficiency requires continuing practice and knowledge transfer
  • Redesign work around complementary human and AI strengths
  • Human oversight should be concentrated where judgment and consequences matter
  • Human oversight must be practical and informed
  • AI workflow redesign should preserve the purposes the human process served

Unlock with Premium — includes all practice exams and the complete study guide.

Scale AI from pilots to enterprise-wide deployments

Read full chapter
  • AI transformation should advance through evidence-driven iterations
  • Useful quick wins create evidence and capabilities for broader adoption
  • A pilot can be valuable even when it does not justify expansion
  • Pilot success must be revalidated in the settings selected for expansion
  • An AI center of excellence should enable business value across teams
  • A CoE can be centralized or federated without abandoning common standards
  • CoE enablement and governance oversight serve different purposes
  • Reusable AI assets need shared quality and contribution practices
  • Scaled AI needs application, business, and AI-quality measures together
  • Success criteria need process baselines and acceptable performance boundaries
  • Feedback needs context and review to drive continuous improvement
  • AI value and return on investment must be tracked after launch
  • Changing inputs and changing business expectations can both erode AI usefulness
  • Feedback should reach the function that can address the diagnosed failure
  • Expert sampling checks live output quality rather than authorizing individual actions
  • Production acceptance requires more than proof that the AI feature works
  • Changes to AI behavior need controlled evaluation and release approval
  • Production handover must establish support ownership and escalation paths
  • Governance evidence must remain current throughout production use
  • Evaluation ground truth must be verified before it can establish output quality
  • Enterprise rollout should stage capabilities after preparing the affected teams
  • Release evaluation methods differ in purpose and user exposure
  • Critical AI-enabled processes need usable fallback arrangements
  • Incident plans must connect safe containment to business recovery requirements
  • High availability sustains service through failures rather than only restoring it afterward
  • Monitoring must connect observations to actionable responses
  • Maintenance and support need repeatable mechanisms as AI usage grows
  • Scaled AI cost oversight needs visibility by use case and business unit

Unlock with Premium — includes all practice exams and the complete study guide.