Share via

Unable to connect to Azure search services from AI Foundry Agent (classic) when search services is behind the firewall

Sandesh Poudel 60 Reputation points Microsoft Employee
2026-04-16T06:34:16.74+00:00

Our Azure Search services has been set to access with selected networks and Trusted azure services are allowed to connect . However, as per documentation, when AI foundry agent uses EntraID, it should be able to connect, but i am receiving following errors;

tool_user_error: Error: search_access_error; Unable to access Azure AI Search index. Please ensure that the project managed identity has Search Index Data Reader and Search Service Contributor roles on the Search resource, and the Search resource allows role-based access. () Request is denied as the source is not allowed by applicable rules. The service is set 'publicNetworkAccess: Enabled'. Please review all service's network security settings to ensure the client is allowed.
User's image

Foundry agent has been set up with Standard Agent service network injection and setting up Private endpoint connections to search services results following error:

Can not connect to host, Please ensure Azure AI search connection has the correct endpoint , DNS server answer with no data
User's image

So, now my question is - How can i connect the standard agent from Foundry to Search services when search service is not configured for all networks ?
thanks

Azure AI Search
Azure AI Search

An Azure search service with built-in artificial intelligence capabilities that enrich information to help identify and explore relevant content at scale.


Answer accepted by question author
  1. Praneeth Maddali 8,445 Reputation points Microsoft External Staff Moderator
    2026-04-16T14:41:29.98+00:00

    Hi Sandesh Poudel

    As discussed offline, we have shared the recommended solution and configuration steps with you. Since the guidance has been confirmed, we are proceeding with case closure.

    0 comments No comments

0 additional answers

Sort by: Most helpful

Your answer

Answers can be marked as 'Accepted' by the question author and 'Recommended' by moderators, which helps users know the answer solved the author's problem.