# What happened to Microsoft Applied Skills?

Microsoft Applied Skills launched as a bold bet: free, hands-on, lab-based credentials that prove you can do the work, not just answer questions about it. I bought in early and earned 36 of them. Lately, though, the feed has gone quiet. This is my attempt to work out why!

* * *

## Where I'm coming from

I have passed and earned 36 Applied Skills credentials so far, so I have seen the program from the inside: the labs, the grading, the retakes and the occasional broken environment.

I have also been interviewed by Microsoft Learn about Applied Skills.

%[https://dimitaronai.com/microsoft-learn-interview] 

I am a fan of the idea, which is why the silence bothers me.

* * *

## What I'm seeing

Two things stand out, coming from what I notice day to day.

First, my network has stopped talking about Applied Skills. People used to post a freshly earned credential every week. I cannot remember the last time I saw one pass through my feed.

Second, I have not seen many new credentials added. I've even forgot which one was the latest...

The retirement record points the same way. Microsoft constantly retires credentials. One of them, Create an AI Agent, was announced around February 2026 and retired on May 28, 2026: about three months. Microsoft does not publish reasons for retirements, so the cause remains a guess.

To be fair, Microsoft has not walked away. Its [credentials page](https://learn.microsoft.com/credentials) still promotes Applied Skills and highlights new business-focused credentials aimed at non-technical roles. So this is not a shutdown. The question is whether the program has lost momentum with the people who made it visible in the first place.

* * *

## Was it rushed?

I see three possible friction points. They are hypotheses, not findings.

**The launch pace.** Applied Skills grew fast, and retirements started early. A learner who prepares for a credential and finds it gone will not come back easily.

**The bugs.** Having the tasks not load, not being able to login to the virtual machine, or a session just ending at random. These are only some of the issues I have faced. This matters more than it looks, because Microsoft's policy says a lab that is cancelled or ends abruptly [is scored and counts as an attempt](https://learn.microsoft.com/en-us/credentials/support/retake-policy). A platform fault can therefore cost you three days.

**The 72-hour wait.** The retake rule for Applied Skills is simple, and the comparison with exams is mixed:

| Credential | First retake | Later retakes | Cost |
| --- | --- | --- | --- |
| Applied Skills lab | After 72 hours | After 72 hours each time | Free |
| Role-based exam | After 24 hours | After 14 days, up to 5 in 12 months | Paid, if applicable |

The wait is shorter than the exam rule for later attempts and the lab is free, so on paper it is not harsh. In practice, though, labs are long, and a three-day gap after a grading glitch breaks momentum. I suspect it hurts enthusiasm more than the numbers suggest.

* * *

## Does AI move too fast for credentials?

This is the explanation I find most convincing, and it may also be the most uncomfortable.

Microsoft's own [retirement page](https://learn.microsoft.com/credentials/support/credential-retirement) says the company reviews its credentials because technology changes faster every year and retires those that stop being relevant. That is a reasonable policy. It also exposes a structural problem: an Applied Skill tests a specific task in a specific product, and AI products shift under the learner's feet. A credential built on a tool that is renamed, merged or redesigned within a few months has a short shelf life.

The retirement record fits that story.

There is a second-order effect for the community. If people expect a credential to be retired soon, the incentive to earn and post it drops. Why advertise a skill that the program itself may discard next quarter? Passed credentials do remain on your transcript, though, so the effort is not lost. That is worth saying to anyone who feels discouraged.

The counter-argument is that not everything is moving. Applied Skills such as Azure AI Language lasted almost three years (now retired too), so slower-moving topics still hold value. If I am right, the lesson is that AI-era credentials should be updated continuously, not retired and reissued.

* * *

## My take

I do not think one cause explains it. My best reading is that the program was launched with more breadth than it could keep fresh, early retirements taught learners not to invest, and small platform frictions turned a good idea into a chore. AI's pace made all three worse.

I still believe in the format. A lab that proves you can do something is worth more than a multiple-choice score. Here is what would bring me, and I suspect others, back:

*   Do not count a platform failure as an attempt and shorten the wait after one.
    
*   Publish simple program numbers, such as credentials earned per quarter, so the community can judge momentum for itself.
    
*   Treat AI credentials as living ones: update the lab when the product changes instead of retiring it.
    

* * *