DP-600, DP-700 and DP-800 Updated (19 October 2026): What Changed

Microsoft's change logs only say which areas changed and mark them "minor". Here's what actually moved.

DP-800: SQL AI Developer Associate

Compared with the outline from 12 March 2026:

• Change event streaming (CES) was removed from the list of ways to keep embeddings up to date. The list is now table triggers, Change Tracking, Azure Functions with SQL trigger binding, Azure Logic Apps, CDC and Microsoft Foundry. Microsoft doesn't say why. CES is still a preview feature, and Microsoft's study guides say questions mostly cover generally available features.
• "ANN and ENN" became "ANN and KNN" in "Choose between using ANN and KNN for vector search". This is a change of name, not of concept: Microsoft's own documentation calls exact vector search both "exact nearest neighbor (ENN)" and "k-nearest neighbor (kNN)", and kNN is the term it mostly uses. The choice is still approximate search with a vector index, or exact search across every vector.
• Wording only: "column store indexes" is now "columnstore indexes", and "Microsoft Copilot in Fabric" is now "Copilot in Fabric".

Weights are unchanged: Design and develop database solutions 35–40%, Implement AI capabilities in database solutions 25–30%, and Secure, optimize, and deploy database solutions 35–40%.

DP-700: Fabric Data Engineer Associate

Compared with the outline from 21 July 2026:

• "Process data by using Eventstreams" is now "…by using Eventstream".
• "Optimize Eventstreams and Eventhouses" is now "Optimize Eventstream and Eventhouse".

That's the naming catching up with the product. Microsoft's change log also marks "Configure Microsoft Fabric workspace settings" as changed, but its four items (Spark, domain, OneLake and Apache Airflow workspace settings) read exactly the same before and after. Weights are unchanged at 30–35% for each of the three areas.

DP-600: Fabric Analytics Engineer Associate

Compared with the outline from 21 July 2026, the only change is capitalisation: "Visual Query Editor" is now "Visual query editor". Weights are unchanged: Prepare data 45–50%, Implement and manage semantic models 25–30%, and Maintain a data analytics solution 25–30%.

What to do about it

DP-600 or DP-700: nothing. Keep going.

DP-800: don't spend time on CES as a way to maintain embeddings. Make sure you can explain when exact (kNN) vector search is the right choice and when an approximate (ANN) vector index is.

How we checked

We compared each current study guide (DP-600, DP-700, DP-800) with the version in force before 19 October, item by item. The CES and vector-search details come from Microsoft's SQL documentation on change event streaming and vector search.