Searching for the AI-102 exam? It no longer exists. Microsoft retired it on June 30, 2026, and AI-103 is the exam that took its place.
Quick Answer
- • AI-102 retired June 30, 2026 (11:59 PM Central Standard Time) and can no longer be taken
- • Replacement: AI-103, Microsoft Certified: Azure AI Apps and Agents Developer Associate
- • If you earned Azure AI Engineer Associate before retirement, it stays valid for its normal one-year period
- • AI-103 is built around Python and Microsoft Foundry, with much more weight on generative AI and agents
What happened to AI-102
AI-102, Designing and Implementing a Microsoft Azure AI Solution, was the exam behind the Azure AI Engineer Associate certification. Microsoft's own study guide now carries a retirement notice, and the exam appears on the official recently retired exams list with a June 30, 2026 retirement date.
You will still see AI-102 in search results, courses and old practice tests. None of them lead to a bookable exam any more.
What replaced it: AI-103
AI-103 leads to Microsoft Certified: Azure AI Apps and Agents Developer Associate. Microsoft describes the candidate as an Azure AI engineer who builds, manages and deploys agents and AI solutions on Microsoft Foundry, with Python experience. The responsibilities listed are:
- • Planning and managing Azure AI solutions
- • Implementing generative AI and agentic solutions
- • Implementing computer vision solutions
- • Implementing text analysis solutions
- • Implementing information extraction solutions
For the full domain weights and a week-by-week plan, see the AI-103 study guide.
AI-102 vs AI-103: what changed
| Area | AI-102 (retired) | AI-103 |
|---|---|---|
| Credential | Azure AI Engineer Associate | Azure AI Apps and Agents Developer Associate |
| Status | Retired June 30, 2026 | Current |
| Platform focus | Azure AI services and Microsoft Foundry | Microsoft Foundry, built with Python |
| Generative AI and agents | 15–20% generative AI, 5–10% agents (Dec 2025 outline) | Largest domain (30–35%), including agents |
| Vision, language, extraction | Three separate domains, including custom vision, language understanding, question answering and knowledge mining | Vision, text analysis and information extraction, each 10–15% |
| Passing score | 700/1000 | 700/1000 |
The practical shift: AI-102 was a tour of Azure AI services. AI-103 is a developer exam about building apps and agents, so hands-on Foundry and Python time matters more than memorising service names.
What happens to your Azure AI Engineer certification
Microsoft states that if you completed the Azure AI Engineer Associate certification before it retired, it remains valid for one year. That is the usual validity period for Associate certifications, and it is counted from the date you earned or last renewed it.
Open your Microsoft Learn certification dashboard to see your exact expiry date and whether the renewal assessment is still offered. If it is not, AI-103 is the route to a current Azure AI credential.
Which exam should you take?
Take AI-103 if you:
- • Were planning to take AI-102 and had not yet passed it
- • Hold AI-102 and want a current Azure AI credential after it expires
- • Build with Python on Foundry, Azure OpenAI or agents
Look at AI-200 or AI-900 instead if you:
- • Build the containers, data services and messaging behind AI apps: see AI-103 vs AI-200
- • Are new to AI on Azure: start with the AI-900 study guide
How to prepare for AI-103 coming from AI-102
- Skim the AI-103 skills outline and mark what is new: agents, Foundry-first workflows, content understanding.
- Keep your AI-102 strengths for computer vision, text analysis and document intelligence, but check each against the AI-103 outline.
- Spend the most time on generative AI and agents, the largest domain.
- Build one small agent and one RAG app on Foundry in Python before the exam.
- Finish with timed practice questions to find weak domains.
Start with the free AI-103 practice questions, or see the full AI-103 exam page.
Practice AI-103 with exam-style questions
40 free questions with explanations and Microsoft Learn references. No credit card required.
Start Free AI-103 Practice →Frequently asked questions
Is the AI-102 exam retired?
Yes. Microsoft retired Exam AI-102 (Designing and Implementing a Microsoft Azure AI Solution) on June 30, 2026, at 11:59 PM Central Standard Time. It can no longer be taken.
What replaced AI-102?
AI-103 is the replacement. It leads to the Microsoft Certified: Azure AI Apps and Agents Developer Associate certification, which centers on building and deploying AI solutions and agents with Python and Microsoft Foundry.
Is my Azure AI Engineer Associate certification still valid?
Microsoft states that if you completed the Azure AI Engineer Associate certification before it retired, it remains valid for one year, the normal validity period. Check your Microsoft Learn profile for the exact expiry date and whether a renewal assessment is still offered to you.
Do I have to take AI-103 if I already passed AI-102?
No. Your existing certification stays on your transcript until it expires. AI-103 is the way to hold a current Azure AI credential after that, and it is the exam to take if you have not earned the AI-102 certification yet.
How much does AI-103 cost?
AI-103 is an Associate-level exam at the standard US$165 exam fee. Pricing varies by country, so confirm it when you schedule.
Can I still use AI-102 study material?
Partly. Shared topics such as responsible AI, computer vision, language and document intelligence still help, but AI-103 puts far more weight on generative AI, agents and Microsoft Foundry, so AI-102 notes alone will leave gaps.