Public Preview: Evaluations for indirect prompt injection attacks in Microsoft Foundry

Microsoft announced a preview of this Microsoft Foundry update in October 2024.

In preview1 version

Facts

Microsoft id
xpia-evaluations-public-preview
Status
In preview
Products
Microsoft Foundry
Update type
Features
Product categories
AI + machine learning
Preview
October 2024
First captured by us
(from our first import of this list)
Microsoft last modified
Source
Azure Updates · open the original at Microsoft

Microsoft says

Risk and safety evaluations for indirect prompt injection attacks are now available in public preview, accessible through Azure AI Foundry UI and SDK experiences.

Azure Updates,

Read Microsoft's full post

What Microsoft changed

One version so far. If Microsoft edits it, each change shows here with what was removed and added.

  1. First version we captured ·

    As first seen: In preview
    Status
    In preview

    Seen by us · Microsoft's timestamp · Microsoft's page

Last updated