AI-102 and Azure AI Engineer Associate retired on June 30, 2026. To plan further certification study, compare the current Azure AI Apps and Agents Developer Associate path and its AI-103 requirements. Check your transcript expiry if you hold the old credential, and reuse material only where it matches the current skills and documentation.
What changed after AI-102 retired?
Exam AI-102 and the Azure AI Engineer Associate credential are retired as of June 30, 2026. You cannot take the exam, earn the credential, or renew it. Microsoft's July 2026 partner announcement lists Azure AI Apps and Agents Developer Associate as the new certification for this role.
The credential page states that Azure AI Engineer Associate and its renewal assessment are retired. The archived exam guide records AI-102 retirement on June 30, 2026, at 11:59 PM Central Standard Time.
Microsoft’s July Partner Center announcement lists Azure AI Apps and Agents Developer Associate in the replacement column for the old credential. That establishes the named relationship; personal certification requirements come from the current credential page, which identifies AI-103 as its required exam.
The retirement notice concerns the credential and assessment. It does not establish the current availability of every service from the old syllabus. Check current product documentation when reusing older instructions.
Sources: Microsoft Certified: Azure AI Engineer Associate - Certifications | Microsoft Learn — Credential overview, retirement warning, historical exam details; Study guide for Exam AI-102: Designing and Implementing a Microsoft Azure AI Solution | Microsoft Learn — Retirement warning, skills measured as of December 23, 2025; Microsoft Certified: Azure AI Apps and Agents Developer Associate - Certifications | Microsoft Learn — Current credential overview, exam duration and policy; Study guide for Exam AI-103: Developing AI Apps and Agents on Azure | Microsoft Learn — Skills measured as of April 16, 2026; July 2026 announcements - Partner Center announcements | Microsoft Learn — Retiring certification and replacement listing
What happens to an existing credential?
An unexpired Azure AI Engineer Associate credential stays active until its own expiry date. Its renewal assessment is no longer available.
Microsoft’s current policy gives role-based certifications one year of validity after requirements are completed. Your record may also reflect a renewal completed before retirement, so check the actual transcript expiry instead of adding one year to the original award date.
Retirement did not cancel unexpired credentials early. A new credential has its own requirements; do not assume that holding AI-102 completes AI-103.
Sources: Microsoft Certified: Azure AI Engineer Associate - Certifications | Microsoft Learn — Credential overview, retirement warning, historical exam details; Credential expiration policies | Microsoft Learn — Role-based certification expiry and retired credential transcript handling; Microsoft Certified: Azure AI Apps and Agents Developer Associate - Certifications | Microsoft Learn — Current credential overview, exam duration and policy; Microsoft Certification Renewal | Microsoft Learn — Six-month renewal window
Which older topics can help with further study?
The historical English skills list dated December 23, 2025 covered six areas across planning, generative and agentic work, vision, language, and information extraction.
The retired weights were planning and management 20–25%, generative AI 15–20%, agentic solutions 5–10%, vision 10–15%, natural language processing 15–20%, and knowledge mining and information extraction 15–20%. These are historical ranges, not AI-103 weights.
Existing explanations of model selection, retrieval, prompting, evaluation, search, extraction, and responsible AI can support further learning where they match current tasks. A shared concept or service name does not guarantee that an old API, configuration, or example still applies.
The historical profile named Python or C# experience. The current AI-103 profile expects Python application development and familiarity with AI and Azure. Check that expectation when coming from C#-based preparation.
Sources: Microsoft Certified: Azure AI Engineer Associate - Certifications | Microsoft Learn — Credential overview, retirement warning, historical exam details; Study guide for Exam AI-102: Designing and Implementing a Microsoft Azure AI Solution | Microsoft Learn — Retirement warning, skills measured as of December 23, 2025; Microsoft Certified: Azure AI Apps and Agents Developer Associate - Certifications | Microsoft Learn — Current credential overview, exam duration and policy; Study guide for Exam AI-103: Developing AI Apps and Agents on Azure | Microsoft Learn — Skills measured as of April 16, 2026
What to study for the current AI-103 path
Use the AI-103 English skills list dated April 16, 2026. Its five areas include a combined generative and agentic solutions area.
Read the current audience profile and detailed tasks before choosing your preparation plan. The percentage ranges below do not prescribe exact question counts or study hours.
Compare each old resource with the current task it is meant to teach. Check service names, supported features, SDK usage, and deployment requirements in current documentation. Do not assume adding a branding update or one agent example makes a whole old course current.
| Current AI-103 area | Weight range | Topics and suggested practice |
|---|---|---|
| Plan and manage an Azure AI solution | 25–30% | Model and service selection, identity and network controls, costs, observability, and responsible AI. Explain the requirements and evidence needed for a fictional design choice. |
| Generative and agentic solutions | 30–35% | Foundry applications, retrieval-augmented generation (RAG), single and multiple agents, tools, memory, and evaluation. Practice checking an answer against supplied information and use official tasks for implementation work. |
| Computer vision solutions | 10–15% | Image and video understanding and generation. Compare listed capabilities with a stated input and desired output using current documentation. |
| Text analysis solutions | 10–15% | Language, translation, speech, and related Foundry capabilities. Review the listed tasks and explain how inputs and outputs differ. |
| Information extraction solutions | 10–15% | Indexing and search, optical character recognition, and Content Understanding. Trace how a small document example is processed and how extracted information would be checked. |
Sources: Study guide for Exam AI-102: Designing and Implementing a Microsoft Azure AI Solution | Microsoft Learn — Retirement warning, skills measured as of December 23, 2025; Microsoft Certified: Azure AI Apps and Agents Developer Associate - Certifications | Microsoft Learn — Current credential overview, exam duration and policy; Study guide for Exam AI-103: Developing AI Apps and Agents on Azure | Microsoft Learn — Skills measured as of April 16, 2026; July 2026 announcements - Partner Center announcements | Microsoft Learn — Retiring certification and replacement listing
Practice: does each answer follow the supplied source?
For this exercise, apply a stated rule: use only the facts in the notice and acknowledge information that it does not provide.
The current AI-103 outline includes evaluation of generative and agentic solutions. This original exercise practices one part of evaluation with a fictional notice and three candidate replies.
The source-only rule is an explicit premise for this exercise, not a universal definition of every AI application. The fixture does not use actual venue hours or real model outputs and does not demonstrate overall model accuracy.
Practice exercise
A venue's retrieved notice reads exactly: 'The visitor desk opens Monday and Wednesday from 09:00 to 17:00. It is closed on Tuesday.' The venue is fictional. Under the rule above, evaluate these three candidate replies: (A) To 'Is the desk open on Tuesday?' the reply is 'It is closed on Tuesday.' (B) To 'What are the desk's hours on Wednesday?' the reply is 'It is open from 09:00 to 18:00.' (C) To 'Is the desk open on Sunday?' the reply is 'The notice does not state Sunday hours.' Which replies follow the rule?
Show answer
Replies A and C follow the rule. Reply B does not.
A matches the notice’s statement that Tuesday is closed. B contradicts the stated Wednesday closing time: it should say 17:00, not 18:00. C correctly identifies that Sunday hours are not supplied. Under this exercise’s rule, an unsupported assertion about Sunday would fail even if it happened to be true elsewhere. This checks source adherence for these three replies, not all aspects of an AI system’s quality or safety.
Your next-study checklist
Use the current AI-103 skills to decide what to keep, what to update, and what to practice.
- Check your actual transcript expiry if you hold Azure AI Engineer Associate.
- Compare the current AI-103 audience profile and role with your goals.
- Map each of the five current areas to material you understand and gaps needing study.
- Review Python expectations if your older preparation used C#.
- Work the source-checking exercise and explain each reply’s result.
- Use current official learning tasks for Foundry applications, agents, retrieval, and evaluation, checking resource requirements before starting.
- Use the signed-in practice assessment and sandbox, and check current booking details.
Sources: Microsoft Certified: Azure AI Engineer Associate - Certifications | Microsoft Learn — Credential overview, retirement warning, historical exam details; Credential expiration policies | Microsoft Learn — Role-based certification expiry and retired credential transcript handling; Microsoft Certified: Azure AI Apps and Agents Developer Associate - Certifications | Microsoft Learn — Current credential overview, exam duration and policy; Study guide for Exam AI-103: Developing AI Apps and Agents on Azure | Microsoft Learn — Skills measured as of April 16, 2026
Current preparation resources and exam details
AI-103 is the current required exam for Azure AI Apps and Agents Developer Associate. Its credential page gives 120 minutes and links current preparation resources.
Microsoft links a practice assessment through AI Skills Navigator, which requires sign-in, and provides an exam sandbox. Use practice explanations to identify gaps and the sandbox to learn the interface; neither the exercise below nor a practice score establishes complete readiness.
Use the current credential page for scheduling and the price in your testing region. Microsoft recommends a personal account so records remain accessible if you leave a school or employer. These are AI-103 preparation steps, not a way to book retired AI-102.
Sources: Microsoft Certified: Azure AI Apps and Agents Developer Associate - Certifications | Microsoft Learn — Current credential overview, exam duration and policy; Study guide for Exam AI-103: Developing AI Apps and Agents on Azure | Microsoft Learn — Skills measured as of April 16, 2026
Sources
Facts checked against retrieved official Microsoft pages:
- Microsoft Certified: Azure AI Engineer Associate - Certifications | Microsoft Learn
- Study guide for Exam AI-102: Designing and Implementing a Microsoft Azure AI Solution | Microsoft Learn
- Credential expiration policies | Microsoft Learn
- Microsoft Certified: Azure AI Apps and Agents Developer Associate - Certifications | Microsoft Learn
- Study guide for Exam AI-103: Developing AI Apps and Agents on Azure | Microsoft Learn
- July 2026 announcements - Partner Center announcements | Microsoft Learn
- Microsoft Certification Renewal | Microsoft Learn