- What "Training" Means for This Exam
- The Six Domains as a Training Map
- Which Outline to Train Against
- How the Exam Format Shapes Your Training
- The 150-Hour Readiness Profile
- Sequencing the Domains Across Your Weeks
- Hands-On Practice That Maps to Objectives
- Registration, Vouchers, and Delivery
- After You Pass: Validity and Career Use
- Frequently Asked Questions
- The current English exam objectives have six domains, from AI Problem Definition through Maintaining and Monitoring AI in Production.
- Pearson's resources page still links a legacy five-domain outline; train against the revised six-domain document instead.
- Expect 45-50 questions in a 50-minute timed limit, with a scaled passing score of 700 out of 1000.
- The program suggests roughly 150 hours of readiness; this is a profile, not a logged-hours requirement.
What "Training" Means for This Exam
The Information Technology Specialist - Artificial Intelligence Certification (ITS-AI) is issued by Pearson and delivered through Certiport, a Pearson VUE business. It is a computer-based, proctored knowledge assessment. That distinction matters for how you train: you are not building a portfolio or defending a project, you are preparing to answer a timed set of objective questions about the full lifecycle of an AI solution.
Good ITS-AI training therefore has two jobs. First, it must cover the conceptual territory in the official exam objectives, which run from deciding whether a problem is suited to AI all the way to retiring a model. Second, it must teach you to recognize how that territory is tested, so that you can move quickly through a 50-minute exam without second-guessing familiar material. If you are new to the credential itself, the overview in What Is ITS-AI Certification? gives useful background before you commit to a plan.
The Six Domains as a Training Map
The revised objectives document, copyright 2026, lists six domains. No official percentage weights are published, so you should not assume any domain is "smaller" and safe to skim. The six domain lines are the only structural guidance you get. For a deeper walkthrough of each area, see ITS-AI Exam Domains 2026: Complete Guide to All 6 Content Areas.
Domain 1: AI Problem Definition
Training here is about judgment before technology. You need to understand how to frame a business need as an AI problem and decide whether AI is the right tool at all.
- Feasibility: whether the data, resources, and problem shape support an AI approach
- Human interaction considerations in how an AI system will be used
- Framing the problem so success can be evaluated later
Domain 2: Data Collection and Transformation
Data work underpins every later domain. Expect to reason about sourcing, quality, and preparation of data before any model is trained.
- Collecting and assessing data for suitability
- Cleaning and transforming raw data into usable form
- Recognizing how data problems propagate into model problems
Domain 3: Feature Engineering
In the current outline, feature engineering stands as its own domain rather than being folded into data preparation. Train it as a distinct skill.
- Creating, selecting, and transforming features
- Understanding why feature choices influence model behavior
- Connecting feature decisions to the data transformations that precede them
Domain 4: AI Algorithms and Models
This domain covers model families and the means of judging them. Current coverage includes natural language processing, explainability, and model-evaluation metrics.
- Matching algorithm types to problem types
- Natural language processing concepts
- Explainability: why a model produced an output and how to communicate it
- Choosing and interpreting evaluation metrics
Domain 5: Application Integrity and Deployment
Domain 5 uses the exact wording preserved from the revised PDF. Application integration remains objective 5.1, and the domain extends into the operational practices that make a deployment trustworthy.
- Integrating AI capabilities into applications
- CI/CD practices for AI solutions
- Security considerations for deployed AI
Domain 6: Maintaining and Monitoring AI in Production
The final domain treats an AI system as something that must be watched and eventually retired. The final objective is 6.5.
- Monitoring performance after release
- Drift: recognizing when data or behavior changes degrade results
- Decommissioning a model or system responsibly
Which Outline to Train Against
One of the most practical training decisions is choosing the right source document. Pearson's general IT Specialist resources page still links a legacy copyright-2025 outline with five domains. The revised copyright-2026 objectives document, which Certiport links from its exam page, has six. They are not interchangeable.
| Item | Legacy outline | Revised outline |
|---|---|---|
| Copyright year | 2025 | 2026 |
| Number of domains | Five | Six |
| Status for English coverage | Must not be treated as complete current coverage | The current reference for the English exam |
| Feature engineering, explainability, CI/CD, drift, decommissioning | Not presented as separate current coverage | Covered as distinct topics |
The English update took effect June 30, 2026, as listed in Certiport's exam content updates archive, following an announcement dated April 29, 2026. Rollout dates for other languages differ, so non-English candidates should check their own language listing rather than assume the same date applies.
How the Exam Format Shapes Your Training
Pearson's MindhubPro listing estimates 45 to 50 questions with a 50-minute maximum. That is a variable range, not a fixed form count, so do not train for exactly one number. The 50 minutes is timed exam time and excludes administration and tutorials, which means the clock does not run while you read the on-screen instructions.
The passing score is 700 on a 1-1000 scale. This is a scaled score, not a raw 70 percent, so you cannot convert it into a simple "number of questions I can miss." Certiport does not publish pass rates, which is worth keeping in mind whenever you see a specific figure quoted elsewhere; our discussion in ITS-AI Pass Rate 2026: What the Data Shows explains how to treat that gap. For scoring detail, see ITS-AI Passing Score 2026: Exactly What You Need to Pass.
Pacing implications
With roughly one minute per question available, training should build fast recognition of concepts rather than slow derivation. For a conceptual exam like this one, that means you should be able to distinguish similar ideas on sight: for example, data drift versus a model that was never well evaluated, or an explainability need versus a security need. Practice by explaining each distinction aloud in one or two sentences.
Interface familiarity
The IT Specialist Exam Tutorial (copyright 2021) describes answer-option navigation. Its generic 36-40 question estimate is older than the exam-specific listing used here, so rely on the 45-50 range for planning but use the tutorial to become comfortable with how options are selected and reviewed.
The 150-Hour Readiness Profile
Certiport's program suggests approximately 150 hours of preparation, and the objective profiles describe candidates with at least 150 hours of relevant experience or study. These are readiness profiles, not a verified logged-hours application requirement. Nobody checks a timesheet when you register. There is also no degree requirement beyond knowledge of the objective domains, a point expanded in ITS-AI Requirements 2026: Eligibility, Prerequisites & How to Qualify.
Treat 150 hours as a calibration tool. If you already work with data pipelines or have shipped a machine learning feature, you may need far less time on Domains 2 and 5 and more on the conceptual vocabulary of Domains 1 and 4. If you are coming from general IT support, the opposite may hold. How demanding this feels in practice is covered in How Hard Is the ITS-AI Exam? Complete Difficulty Guide 2026.
Key Takeaway
Use the 150-hour figure to size your plan, then adjust by domain based on a diagnostic. Take a practice set early, tag every miss to one of the six domains, and spend your hours where the misses cluster rather than splitting time evenly.
Sequencing the Domains Across Your Weeks
The domains follow the lifecycle of an AI project, and that order is also a sensible training order, because each domain assumes vocabulary from the one before it. Below is one way to map them to a six-week plan. Adjust the pacing to your own diagnostic results, and see the ITS-AI Study Guide 2026 for a broader approach.
AI Problem Definition
- Learn to judge feasibility and frame problems
- Cover human interaction considerations
- Take a baseline diagnostic across all six domains
Data Collection and Transformation
- Study sourcing, quality, and cleaning
- Work through transformation examples by hand
Feature Engineering
- Practice creating and selecting features
- Tie feature choices back to Week 2 data issues
AI Algorithms and Models
- Match algorithm types to problem types
- Cover NLP, explainability, and evaluation metrics
- This is the densest domain, so consider extra days
Application Integrity and Deployment
- Review integration, CI/CD, and security
- Link deployment choices to the models from Week 4
Monitoring, Then Full Review
- Cover drift, monitoring, and decommissioning
- Run timed mixed sets of 45-50 questions in 50 minutes
- Revisit your weakest domains from the diagnostic
The reason to front-load Problem Definition is that its vocabulary of feasibility, success criteria, and human interaction reappears when you later study evaluation metrics and monitoring. Questions in Domains 4 and 6 often make more sense if you remember what "success" was supposed to mean in Domain 1.
Hands-On Practice That Maps to Objectives
Because the exam is knowledge-based, you do not need a production environment. Small, targeted exercises are enough to make the concepts concrete.
- Data and features (Domains 2-3): take a small public dataset, document its quality problems, clean it, and then create at least two engineered features. Write down why each feature might help a model.
- Models and metrics (Domain 4): train a simple model, compute more than one evaluation metric, and note where the metrics disagree about whether the model is good.
- Explainability (Domain 4): take one prediction and write a plain-language explanation a non-technical stakeholder could follow.
- Deployment and security (Domain 5): sketch how a model would be integrated into an application, what an automated build-and-release pipeline would check, and where the security risks sit.
- Monitoring (Domain 6): define what you would measure after release, what would signal drift, and what conditions would trigger retirement of the model.
A short written note for each exercise, mapped to a numbered objective, doubles as a revision sheet. When you are close to your exam date, condense those notes using the approach in the ITS-AI Cheat Sheet 2026.
Registration, Vouchers, and Delivery
Training ends at a booking screen, and the mechanics there can surprise candidates. Two voucher routes appear in the available pricing information.
| Option | Price shown | Key conditions |
|---|---|---|
| Pearson MindhubPro exam voucher | USD 127 web/member pricing | Country-restricted and unavailable in some listed markets; expires 12 months after purchase |
| CATC voucher with retake | USD 133 | US-only; proctoring may cost extra; not usable at Pearson VUE centers or through OnVUE |
Voucher expiry is a separate matter from credential validity. The 12-month window applies to when you must use the voucher, not to how long your certification lasts. A full cost breakdown, including how to think about retakes, is in ITS-AI Certification Cost 2026: Complete Pricing Breakdown.
Choosing a delivery channel
You can test through Pearson VUE test centers or through OnVUE online proctoring, or through the separate Certiport Authorized Testing Center (CATC) channel. OnVUE requires a system check and a live webcam proctor, so run the check well before exam day. At a CATC, expect photo and signature ID, a non-disclosure agreement, parental consent if you are under 18, and a rule against personal paper or communication devices. Scheduling considerations are discussed in ITS-AI Exam Dates 2026: Testing Windows, Deadlines & Scheduling.
After You Pass: Validity and Career Use
For awards earned on or after 2025-06-18, the certification is valid for five years. Earlier awards remain nonexpiring. Recertification requires purchasing a new voucher and passing the current exam, so a future renewal will be tested against whatever objectives are current at that time rather than the ones you studied.
The credential is an entry-level signal of AI lifecycle literacy rather than a substitute for hands-on engineering experience. It tends to be most useful for candidates who want to show structured understanding of how AI projects are scoped, built, deployed, and maintained. For how it fits into the job market, read ITS-AI Jobs, and for the financial side, see the ITS-AI Salary Guide 2026 and Is the ITS-AI Certification Worth It? Complete ROI Analysis 2026. When you are ready to test your recall under timed conditions, the ITS-AI Exam Prep practice exams are built to mirror the six current domains.
Frequently Asked Questions
The revised exam objectives list six domains: AI Problem Definition, Data Collection and Transformation, Feature Engineering, AI Algorithms and Models, Application Integrity and Deployment, and Maintaining and Monitoring AI in Production. Pearson has not published official percentage weights for them.
No. That legacy copyright-2025 outline is still linked from Pearson's general IT Specialist resources page, but it should not be treated as complete current English coverage. Train against the revised copyright-2026 six-domain objectives instead.
Pearson's listing estimates 45 to 50 questions with a 50-minute maximum of timed exam time. The passing score is 700 on a 1-1000 scale, which is a scaled score rather than a raw 70 percent requirement.
No. The roughly 150-hour figure is a readiness profile suggested by the program, not a verified logged-hours application requirement, and there is no degree requirement beyond objective-domain knowledge.
Awards earned on or after 2025-06-18 are valid for five years, while earlier awards remain nonexpiring. Recertifying requires a new voucher and passing the current exam. Voucher expiry, which is 12 months after purchase for the MindhubPro voucher, is separate from credential validity.