- GAF here means Generative AI Foundations, a Pearson certification delivered through Certiport.
- The exam has four domains, from generative AI methods to ethics, law, and societal impact.
- You get 50 timed minutes, and a scaled score of 700 on a 1-1000 scale is required to pass.
- U.S. Certiport vouchers cost USD 72 (single) or USD 83 (with retake), valid only at CATC sites.
The Short Answer: What a GAF Is
If you landed here asking "what is a GAF," the answer on this site is specific: GAF stands for Generative AI Foundations, an entry-level certification that tests whether a candidate understands how generative AI works, can write and improve prompts, and can reason about the ethical and legal consequences of using these tools. It is a proctored, computer-based knowledge assessment, not a hands-on lab and not a coding exam.
The acronym is shared by other, unrelated credentials in other industries. Nothing here applies to those. Everything in this article describes Pearson's Generative AI Foundations exam only. If you want the same ground covered from other angles, see our explainers on what GAF is, what GAF stands for, and what GAF certification is.
Who Issues It and Where It Sits
The Generative AI Foundations exam is a Pearson credential. It is developed and administered through Certiport, a Pearson VUE business, and it belongs to the Critical Career Skills (CCS) family of certifications. That family positioning matters because it explains the exam's tone: it is built as a foundational, broadly accessible credential for students, career changers, and working professionals who need verified baseline literacy rather than deep engineering skill.
Pearson publicly announced the certification on October 1, 2024, and the published exam objectives document carries a 2024 copyright. That date identifies the source document rather than a separately verified exam-version number, so treat the objectives as the authoritative content outline and check the issuer's page before you book in case anything has been updated.
The Four Exam Domains
The official objectives are organized into four domains. Important caveat: the published outline lists the domain headings but does not publish percentage weights, so any article claiming exact domain percentages is guessing. Plan to be competent across all four. For a deeper walkthrough, read our complete guide to all four GAF content areas.
Domain 1: Generative AI Methods and Methodologies
This is the conceptual foundation. Expect questions on what generative AI is, how it differs from other kinds of AI, and the general methods behind systems that produce text, images, and other content.
- Distinguish generative models from traditional predictive or rule-based systems
- Recognize the kinds of content generative systems can create and typical use cases
- Understand, at a conceptual level, how models learn from data and why that shapes their outputs
- Know common limitations, such as confident-sounding but incorrect output
Domain 2: Basic Prompt Engineering
The practical core of the exam. You need to know what makes a prompt effective and be able to recognize a strong prompt versus a weak one.
- Components of a well-formed prompt: task, context, constraints, and desired format
- How specificity, role framing, and examples change results
- Matching a prompt to a goal, such as summarizing, drafting, or brainstorming
Domain 3: Prompt Refinement
Writing a first prompt is rarely enough. This domain tests whether you can evaluate an output and improve it through iteration.
- Diagnosing why an output missed the mark: vague instruction, missing context, wrong format
- Revising prompts step by step instead of starting over
- Using follow-up instructions to adjust tone, length, structure, and accuracy
Domain 4: Ethics, Law, and Societal Impact
Many candidates under-prepare here because it feels less technical. It is not a throwaway section; it is one of four named domains.
- Bias in training data and how it can surface in outputs
- Privacy, confidentiality, and what is appropriate to put into a prompt
- Intellectual property and attribution questions around generated content
- Responsible use, transparency, and the wider effect of AI on work and society
What the Exam Looks Like
The test is a proctored, computer-based knowledge assessment. You select answers on screen, can flag questions you are unsure about, and can review before final submission. That flag-and-review workflow is worth practicing, because it lets you bank easy points first and return to the harder items.
| Feature | What the Issuer Specifies |
|---|---|
| Format | Proctored computer-based knowledge assessment |
| Timed duration | 50 minutes (excludes tutorial and administrative seat time) |
| Question count | Sources conflict: the Certiport tutorial specifies 40-45, while the Mindhub voucher page estimates 45-50 |
| Passing score | 700 on a 1-1000 scaled-score system |
| Navigation | Answer selection, flagging, and review before submission |
| Domains | Four, with no published percentage weights |
A 700 on a 1-1000 scale is a scaled score, not a percentage of questions answered correctly. You cannot convert it into "I need X out of Y right." Our passing score breakdown explains how to think about it, and our difficulty guide covers what makes the questions feel easy or tricky.
Vouchers, Fees and Booking Mechanics
This is the area where candidates most often get tripped up, because there are two separate voucher routes and they are not interchangeable.
Certiport route (U.S. pricing)
- Single-exam voucher: USD 72
- Voucher with a retake: USD 83
- A Certiport Authorized Testing Center (CATC) may add its own proctoring charge on top
- These vouchers are valid only for CATC delivery, either in person or remotely, and cannot be redeemed through OnVUE or a Pearson VUE Testing Center
Pearson Mindhub route
- Mindhub voucher: USD 85; Mindhub Pro lists both web and member prices as USD 85
- Mindhub vouchers carry country restrictions and must match the country selected at checkout
Pearson's launch announcement also names an OnVUE delivery route. However, the public Mindhub scheduling link currently redirects to an Adobe scheduling page, so a current GAF-specific OnVUE booking process could not be verified. The practical rule: decide where you will test first, then buy the voucher that matches that channel, and confirm compatibility before paying. For the full cost picture including retake logic, see our GAF certification cost breakdown, and for scheduling specifics see GAF exam dates and scheduling.
Key Takeaway
Do not buy a voucher until you know your delivery channel. A Certiport voucher works only at a CATC, and a Mindhub voucher is tied to the country you select at checkout. Mismatching voucher and channel is the most avoidable mistake in the whole process.
Who Should Sit the Exam
The exam is described as best suited for candidates aged 14 and older. There is no bachelor's degree requirement. Instead, the issuer expects roughly 150 combined hours of instruction and hands-on use of generative AI tools, along with familiarity with productivity applications. Those hours describe how much preparation the issuer assumes, not the length of the exam and not a mandatory course.
Testing rules are enforced through the CATC. Proctoring is required, valid identification is required, and candidates under 18 need parental consent. Books and personal devices must be inaccessible during the exam. Full details are in our GAF requirements guide.
How Long the Credential Lasts
CCS certifications earned on or after June 18, 2025 are valid for five years. Renewal means buying a new voucher and passing the then-current exam. Earlier nonexpiring CCS certifications keep their existing status. Because generative AI changes quickly, a five-year window is a long time in this field; expect the current exam to differ from today's when you renew.
Where the Credential Helps
Generative AI Foundations is a literacy credential, so its value is strongest where employers or schools want proof that a person can use these tools sensibly and safely. That typically means students building a resume, career changers adding a verified AI skill, and professionals in office, marketing, communications, education, and administrative roles where AI-assisted work is becoming routine. It signals baseline competence; it does not replace engineering or data-science credentials.
Be cautious with any specific earnings claims. We do not cite salary figures here because none are published for this credential by the issuer. For a measured discussion of career value, read our ROI analysis, our salary guide, and our overview of GAF-related jobs.
Sequencing Your Preparation by Domain
You do not need an elaborate system. What matters is ordering the domains so that each one builds on the last, and leaving time to practice under timed conditions. Here is a sensible four-week arc for a candidate who already has some hands-on exposure to generative AI tools.
Domain 1: Methods and Methodologies
- Lock down vocabulary and the difference between generative and other AI
- Note limitations such as inaccurate or biased output, which return in Domain 4
Domain 2: Basic Prompt Engineering
- Practice building prompts with task, context, constraints, and format
- Compare weak and strong versions of the same request
Domain 3: Prompt Refinement
- Take mediocre outputs and iterate toward better ones, naming what each revision fixed
- This pairs naturally with Week 2 because refinement only makes sense after you can write a baseline prompt
Domain 4 plus timed practice
- Cover bias, privacy, intellectual property, and responsible use
- Finish with timed practice sets using the flag-and-review workflow
For a fuller plan, use our GAF study guide, keep the cheat sheet handy for final review, and check the training overview if you want structured instruction. When you are ready to test yourself against exam-style questions, try the GAF practice test, and use the full practice question bank to find weak domains before you book.
Frequently Asked Questions
It is the Generative AI Foundations certification from Pearson, delivered through Certiport within the Critical Career Skills program. It validates foundational knowledge of generative AI, prompting, and responsible use.
The timed portion is 50 minutes, excluding the tutorial and administrative time. You need a scaled score of 700 on Certiport's 1-1000 scale, which is not a percentage of correct answers.
Sources disagree. The Certiport tutorial specifies 40-45 questions, while the Mindhub voucher page estimates 45-50. Certiport notes totals differ by exam form, so plan for either range.
U.S. Certiport vouchers are USD 72 for a single exam or USD 83 with a retake, plus any CATC proctoring charge. The separate Mindhub voucher is USD 85 and is subject to country restrictions.
CCS certifications earned on or after June 18, 2025 are valid for five years, and renewal requires a new voucher and passing the current exam. Earlier nonexpiring certifications keep their existing status.