AICP logo
Focused certification exam prep
Start practice

AICP Exam Prerequisites and Eligibility Requirements 2026

TL;DR
  • Completing accredited AICP training with Practical Assignments is a mandatory prerequisite - you cannot register for the exam without it.
  • The exam is 40 multiple-choice questions in 90 minutes, with a 65% passing score, administered by EXIN via the EXIN Anywhere platform or on-site.
  • The EU AI Act text is permitted during the exam; understanding how to navigate it quickly is a core exam skill, not a shortcut.
  • Domain 2 (Articles 8, 9, and 10 analysis) carries the highest weight at 25% - it demands deep technical and legal interpretation, not surface recall.

Who the AICP Certification Is Actually For

The Artificial Intelligence Compliance Professional (AICP) certification is not a general AI literacy badge. It is a specialist credential designed for professionals who will be directly responsible for ensuring that AI systems meet legal, ethical, and governance requirements - particularly under the EU AI Act, which began phased enforcement in 2025 and reaches full application through 2027.

If your role involves any of the following, the AICP is worth evaluating seriously:

  • Data protection and privacy officers who need to extend their compliance mandate into AI system oversight
  • Risk and compliance managers in organizations developing or deploying AI systems in the EU market
  • Legal and regulatory affairs professionals advising on AI Act obligations, prohibited uses, and high-risk AI classification
  • AI product managers and technical program managers responsible for embedding compliance requirements into development lifecycles
  • Auditors and consultants building practice areas around AI governance, ISO/IEC 42001 implementation, or NIST AI RMF alignment

The AICP is notable as the first certification to integrate the EU AI Act, ISO/IEC 42001 (the international standard for AI management systems), and the NIST AI Risk Management Framework into a single lifecycle-based compliance curriculum. That integration is what makes the eligibility question interesting - you need more than general curiosity to succeed here.

Why 2026 Is a Critical Year to Certify: The EU AI Act's obligations for high-risk AI systems come into full effect in 2026, and prohibited AI practices have already been enforceable since February 2025. Organizations are actively building compliance functions right now, and the market for credentialed AICP professionals is expanding in direct proportion to that enforcement timeline.

The Mandatory Prerequisite: Accredited Training with Practical Assignments

This is the non-negotiable eligibility gate: you must complete an accredited AICP training program that includes Practical Assignments before you can sit the EXIN exam. Unlike many professional certifications where training is recommended but exam registration is open to self-study candidates, EXIN has structured the AICP so that accredited training is a hard prerequisite.

What "Accredited Training with Practical Assignments" Actually Means

EXIN authorizes a network of Accredited Training Organizations (ATOs) to deliver AICP preparation. These providers must follow EXIN's official exam literature, which is based on the 2025 curriculum integrating the EU AI Act text, ISO/IEC 42001, and NIST AI RMF. The Practical Assignments component is not optional coursework - it is a structured set of applied exercises designed to demonstrate competency across the compliance lifecycle before you reach the exam.

The total recommended preparation time is approximately 112 hours, broken down as roughly 14 contact hours of formal training plus substantial self-study and assignment work. This is a meaningful commitment, and candidates who treat the practical assignments as checkbox exercises rather than genuine learning opportunities tend to be underprepared for the exam's applied, scenario-based questions.

Training Package Costs: EXIN typically bundles the exam fee within training packages, which range from approximately $800 to $1,700 depending on the accredited provider, format (live instructor-led versus self-paced), and included study materials. There is generally no separate route to purchase an exam voucher without enrolling in an accredited training program.

How to Choose an Accredited Training Provider

When evaluating providers, confirm they are listed on EXIN's official ATO directory and that their package explicitly includes the Practical Assignments component and access to the official EXIN exam literature. Some providers offer additional practice resources - using a dedicated AICP practice test platform alongside your formal training is one of the most effective ways to convert theoretical knowledge into exam-ready application.

While there are no formal educational prerequisites beyond the accredited training requirement, EXIN and training providers consistently recommend that candidates arrive with certain background knowledge to make the most of the 112-hour preparation window.

Background Area Why It Matters for the AICP Urgency if You Lack It
AI fundamentals (ML concepts, model types, AI system lifecycle) Domain 1 and Domain 5 assume you can recognize AI system components without basic definitions consuming exam time High - close gaps before training begins
GDPR familiarity Domain 3 (Privacy, Transparency, Data Governance) builds directly on data subject rights and data minimization principles High - GDPR intersection with the AI Act is frequently tested
ISO 27001 or ISO management system concepts ISO/IEC 42001 borrows heavily from ISO management system structure; familiarity accelerates Domain 5 mastery Medium - helpful but trainable within the 112-hour window
Basic risk management frameworks NIST AI RMF concepts appear in Domains 4 and 5; understanding GOVERN, MAP, MEASURE, MANAGE functions helps Medium - NIST RMF is introduced in training but prior exposure speeds comprehension
EU regulatory environment Understanding how EU directives and regulations differ, and how the AI Act interacts with existing sectoral law, informs Domain 1 questions Low-Medium - covered in training but background helps with context

Exam Mechanics: Format, Fee, and Registration

The AICP exam is administered exclusively by EXIN (Examination Institute for Information Science). EXIN is a Netherlands-based independent examination institute with a long track record in IT and data management certifications, which gives the AICP meaningful institutional credibility in European regulatory contexts.

Exam Format at a Glance

  • Questions: 40 multiple-choice questions
  • Time limit: 90 minutes
  • Passing score: 65%
  • Open book: Yes - the EU AI Act text is permitted during the exam
  • Languages available: English, French, Dutch, Portuguese
  • Certification validity: Lifetime - no mandatory recertification

Testing Delivery Options

EXIN offers two delivery pathways. The EXIN Anywhere platform supports remote proctoring in both live (human proctor monitoring via webcam) and video-recorded formats, giving candidates the flexibility to test from a suitable home or office environment. Alternatively, candidates can sit the exam at an on-site location with one of EXIN's accredited testing partners. Both pathways deliver the same exam and scoring process.

Registration is typically handled through your accredited training provider once you have completed the required training and Practical Assignments. Your provider will issue the necessary confirmation to EXIN, at which point you can schedule your exam slot through the EXIN Anywhere portal.

What the Exam Actually Tests: The Five Domains

The AICP exam is organized across five domains, each weighted differently. Understanding which domains carry the most marks - and what specific knowledge each requires - is essential for targeted preparation. The exam does not reward broad familiarity; it rewards the ability to apply specific provisions of the EU AI Act and supporting frameworks to realistic compliance scenarios.

Domain 1: General Understanding of the EU AI Act (20%)

This domain tests foundational knowledge of the AI Act's structure, scope, and core definitions. Candidates must understand the regulation's risk classification tiers (unacceptable risk, high-risk, limited risk, minimal risk), which AI systems are covered, and how the Act interacts with existing EU law.

  • The definition of an "AI system" under the Act and what falls outside its scope
  • Prohibited AI practices under Article 5 and the rationale for prohibition
  • The role of national competent authorities and market surveillance
  • How the AI Act enforcement timeline phases apply through 2027

Domain 2: In-Depth Analysis of Articles 8, 9, and 10 (25% - highest weight)

This is the most heavily weighted domain and the one where candidates most often lose marks. Articles 8, 9, and 10 cover requirements for high-risk AI systems: conformity requirements, risk management systems, and data governance respectively. Exam questions in this domain require you to interpret specific article provisions and apply them to scenarios involving actual AI systems.

  • Article 9 risk management system requirements: documentation, iterative testing, residual risk controls
  • Article 10 data and data governance: training, validation, testing data quality requirements
  • Article 8 compliance with requirements: how conformity is demonstrated and maintained
  • The interplay between these articles and technical documentation obligations under Article 11

Domain 3: Building Trustworthy AI - Privacy, Transparency, and Data Governance (20%)

This domain bridges the AI Act with GDPR obligations and the broader trustworthy AI framework. Candidates must demonstrate command of transparency obligations, human oversight mechanisms, and data governance practices as they apply across the AI system lifecycle.

  • Transparency and information obligations for limited-risk AI systems (chatbots, deepfakes)
  • Intersection of AI Act data governance requirements with GDPR data minimization and purpose limitation
  • Logging and traceability obligations for high-risk AI systems
  • Fundamental rights impact assessments in AI deployment contexts

Domain 4: Ethical AI Frameworks and Human Rights (15%)

The lowest-weighted domain, but one that requires genuine conceptual fluency. Questions here connect EU ethical AI principles to real governance decisions - not philosophical abstraction, but applied ethics in product and deployment decisions.

  • The seven requirements of the EU's trustworthy AI framework (HLEG guidelines)
  • Human agency and oversight as both an ethical and legal requirement
  • NIST AI RMF's GOVERN function in establishing ethical accountability structures
  • Bias detection, fairness considerations, and non-discrimination requirements under the AI Act

Domain 5: AI Compliance Lifecycle Management and Implementation (20%)

This domain tests whether candidates can actually operationalize compliance - not just describe it. Questions are scenario-based and require understanding how to implement an AI management system aligned with ISO/IEC 42001, integrate NIST AI RMF controls, and maintain compliance across the AI system's operational lifecycle.

  • ISO/IEC 42001 management system structure: context, leadership, planning, support, operation, evaluation, improvement
  • Conformity assessment procedures for high-risk AI: self-assessment versus third-party audit requirements
  • Post-market monitoring obligations and incident reporting under the AI Act
  • Integrating compliance activities into Agile and DevOps AI development workflows

Key Takeaway

Domain 2's 25% weighting means that mastering Articles 8, 9, and 10 in detail is not optional - it is the single most important investment of your study time. Candidates who can apply these three articles to novel scenarios will likely clear the 65% passing threshold; those who only memorize them in isolation may not. Use AICP practice questions specifically targeting Domain 2 scenarios to build this applied fluency.

The Open Book Advantage and What It Means for Preparation

The AICP exam permits candidates to reference the EU AI Act text during the exam. This is a deliberate design choice by EXIN - the exam tests your ability to apply the regulation, not recite it. Understanding exactly what this open book policy means in practice changes how you should prepare.

For a full breakdown of what materials are permitted, how to organize the AI Act text for fast reference, and what the open book policy does and does not help with, see our dedicated guide on the AICP Open Book Policy: What You Can Bring to the Exam.

The critical insight is this: candidates who rely on the AI Act text as a crutch during the exam typically run out of time. With 40 questions and 90 minutes, you have an average of 2 minutes and 15 seconds per question. Navigating an unfamiliar document under time pressure is a liability, not an asset. The open book policy is most valuable for confirming specific article numbers, exact definitions, and precise thresholds - not for reading provisions for the first time.

What to Annotate Before Your Exam: Build a personal index of the AI Act organized by domain topic - Articles 5 (prohibitions), 8-10 (high-risk requirements), 13 (transparency), 17 (quality management), and 72+ (governance and enforcement). Knowing exactly where to look in 20 seconds is the real skill the open book format rewards.

A Domain-Anchored Preparation Timeline

EXIN recommends approximately 112 total preparation hours for the AICP. Given that 14 of those are formal contact hours with your accredited training provider, you have roughly 98 hours of self-directed study to allocate. Here is a domain-weighted approach to that allocation:

Week 1-2

Foundation: Domains 1 and 3

  • Read the EU AI Act in full - don't skim. Annotate as you go for your open book index.
  • Map the AI Act's risk tiers to real AI system examples you know from your industry
  • Review GDPR Articles 5, 25, and 35 to ground Domain 3 data governance content
  • Complete any assigned Practical Assignments from your training provider in this phase
Week 3-4

Deep Dive: Domain 2 (Articles 8, 9, 10)

  • Read Articles 8, 9, and 10 three times - once for comprehension, once for structure, once for application gaps
  • Practice applying Article 9 risk management requirements to three different AI system scenarios (healthcare, credit scoring, biometric identification)
  • Use spaced repetition specifically for Article 10 data governance requirements - these are detail-heavy and scenario questions are unforgiving
  • Run timed Domain 2 practice questions at the AICP practice test platform to identify weak spots
Week 5

Applied Frameworks: Domains 4 and 5

  • Work through ISO/IEC 42001 clause structure - map each clause to a concrete compliance activity
  • Review the NIST AI RMF GOVERN and MAP functions in relation to Domain 4 ethics requirements
  • Domain 4 is lighter in weight (15%) but questions often require integrating ethical reasoning with legal provisions - practice this combination
Week 6-7

Integration and Exam Simulation

  • Complete two full 40-question timed practice exams under open book conditions
  • Review every wrong answer by mapping it to a specific article, clause, or framework element - not just by re-reading the answer
  • Finalize your AI Act annotation index and practice fast-lookup drills
  • Confirm your exam registration and technical setup for EXIN Anywhere if testing remotely

Who Hires AICP-Certified Professionals

The AICP was launched in 2025, making it a very new credential - but the demand it addresses is not new. Organizations across the EU and internationally have been scrambling to build AI governance capacity ahead of the AI Act's enforcement milestones. The AICP certificate provides a structured, verifiable signal of that capability.

Employer demand is currently strongest in several sectors:

  • Financial services and banking: High-risk AI systems in credit scoring, fraud detection, and customer segmentation are directly regulated under the AI Act, and compliance functions in these organizations need staff who understand both the regulatory requirements and the technical systems
  • Healthcare and medtech: AI-assisted diagnostic tools and patient management systems fall under high-risk AI classification, creating demand for compliance professionals who can navigate Article 9 risk management and Article 10 data governance in clinical contexts
  • Management consulting and professional services: The Big Four and specialist AI governance consultancies are building AICP-certified practices to advise clients on AI Act readiness, ISO/IEC 42001 implementation, and conformity assessment
  • Technology vendors and AI platform providers: Companies selling AI systems into the EU market need compliance professionals to manage technical documentation, conformity assessments, and post-market monitoring obligations
  • Public sector organizations: Government bodies deploying AI in law enforcement, social services, or critical infrastructure face some of the most stringent AI Act obligations and are increasingly seeking credentialed expertise

For a comprehensive look at how to position yourself for these roles after earning the credential, the detailed eligibility overview at AICP Exam Prerequisites and Eligibility Requirements 2026 provides additional context on how employers are structuring AI compliance roles around the certification.

Frequently Asked Questions

Can I sit the AICP exam without completing accredited training?

No. Completion of an EXIN-accredited AICP training program that includes Practical Assignments is a mandatory prerequisite. EXIN does not offer a standalone exam-only registration pathway. Your training provider will confirm your eligibility to EXIN before you can schedule an exam slot.

Is there a work experience requirement for the AICP?

EXIN does not publish a formal minimum years of experience requirement for the AICP. However, the recommended background includes AI fundamentals and familiarity with GDPR and ISO 27001 - candidates without any professional exposure to compliance, data governance, or AI development will find the training and exam significantly more challenging. The Practical Assignments component is designed in part to build applied competency that pure theoretical backgrounds lack.

How long does AICP certification last?

The AICP certification is valid for life. EXIN does not currently require recertification or continuing education credits to maintain the credential. However, given that the certification is explicitly tied to the EU AI Act - a regulation that will continue to evolve through implementing acts, delegated acts, and updated guidelines - staying current with regulatory developments is a practical professional necessity even without a formal recertification requirement.

What happens if I fail the exam?

EXIN allows candidates to retake the exam, but retake policies (including waiting periods and any additional fees) are determined at the provider and EXIN level - confirm these specifics with your accredited training provider before you register. Since the exam fee is typically bundled into the training package, clarify whether retake attempts are included in your package or require a separate fee.

Is 65% a difficult passing score to hit for this exam?

EXIN has not publicly disclosed pass rate data for the AICP. What we can say is that 65% on a 40-question exam means you need to answer at least 26 questions correctly. Given that the exam uses scenario-based multiple-choice questions and tests applied knowledge across five domains - with Domain 2 alone accounting for 25% of the total - thorough preparation through accredited training, Practical Assignments, and targeted practice testing is the consistent recommendation from providers and successful candidates alike.

Ready to Start Practicing?

Test your knowledge across all five AICP domains with scenario-based practice questions aligned to the 2025 EXIN exam literature. Identify your gaps in Articles 8, 9, and 10 before exam day - not during it.

Start Free Practice Test

Ready to pass your AICP exam?

Put this into practice with free AICP questions across every exam domain.