AIToday
Large Language ModelsTop Companies' AI MovesTop Companies AIPublished: Nov 22, 2026, 06:30 JST

Google details three auth patterns for StreamAssist API

Send on LINE
Google details three auth patterns for StreamAssist API

3 Key Points

  1. What happened

    Google detailed three keyless authentication patterns for calling the StreamAssist API: Google Identity, Workforce Identity Federation with Microsoft Entra ID or Okta, and Workload Identity Federation for backend bots.

  2. Why it matters

    Enterprises building their own portals now carry the burden of authentication that the out-of-the-box web app handled automatically, so the guide gives them a concrete architecture to follow.

Not sure about something? Ask the AI

Questions and answers are published on this page.

Summaries like this, every morning.

Context & Analysis

The guide draws a clear line between the out-of-the-box Gemini Enterprise web application and a custom-built Customer UI. In the built-in app, identity and access management happen behind the scenes: employees sign in through Google Identity or Workforce Identity Federation, search results respect per-user access controls, and third-party login prompts appear automatically. Once an application calls the StreamAssist API directly, those three authentication flows become the responsibility of the Customer UI and Customer BE.

The document then breaks the problem into three layers. Part 1 covers authenticating to StreamAssist with three keyless patterns: pure Google Identity, Workforce Identity Federation via Google STS, and Workload Identity Federation for backend bots. Part 2 walks through an 11-step 3-Legged OAuth flow for Federated Connectors, including a note that custom domains need a flag enabled so the callback page posts back to the Customer UI. Part 3 addresses custom ADK agents, explaining that root-level OAuth blocks the entire agent upfront and recommending instead wrapping protected tools with GcpAuthProvider and deploying with Agent Identity turned on, which triggers an on-demand popup only when a protected tool is called.

A runnable code blueprint is packaged in the gemini-enterprise-custom-ui-auth GitHub repository, with standalone Python scripts and curl commands for the user and workload auth patterns, a CLI helper and interactive web test UI for the Federated Connector flow, and a multi-agent ADK implementation with deployment script for the on-demand tool OAuth flow.

FAQ
What IAM role is required to call the StreamAssist API?
Every principal calling the StreamAssist API must be granted the Discovery Engine User IAM role on the GCP project. For human user patterns, the user or Workforce Pool group must also have an active Gemini Enterprise user license assigned in the Admin Console.
When should a company use Workload Identity Federation instead of user identity patterns?
Workload Identity Federation is for non-personalized backend bots or shared service portals over shared knowledge bases. It authenticates a machine rather than an individual employee, so all requests appear as the same workload identity.
How are third-party connector tokens stored for a custom UI?
Gemini Enterprise stores per-user refresh tokens inside its own managed Keystore, so Customer BE never has to store third-party credentials in its own database.
Top Companies AIRead Original Article
Send on LINE

AI news that matters for your work, delivered every morning.

Pick your industry and the AI tools you use, and get news related to your work every day.

Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →

Ask AI

Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.

Questions and answers are published on this page.

Related Articles

Next articleAccenture demos AI agents that consult each other on loans