From 451f1e6a2e4764f4a9ac58bef8987bd811d46272 Mon Sep 17 00:00:00 2001
From: Kjell Tore Guttormsen <hello@fromaitochitta.com>
Date: Wed, 24 Jun 2026 12:15:37 +0200
Subject: [PATCH] =?UTF-8?q?docs(ms-ai-architect):=20KB-refresh=20LOW=2026/?=
 =?UTF-8?q?52=20=E2=80=94=20security-hardening-ai-gateway:=20llm-content-s?=
 =?UTF-8?q?afety=20syntaks=20(text-blocklist-ids=E2=86=92blocklists,=20shi?=
 =?UTF-8?q?eld-prompt,=20element-rekkef=C3=B8lge)?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 .../api-management/security-hardening-ai-gateway.md  | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

diff --git a/skills/ms-ai-engineering/references/api-management/security-hardening-ai-gateway.md b/skills/ms-ai-engineering/references/api-management/security-hardening-ai-gateway.md
index da8ba2b..117d408 100644
--- a/skills/ms-ai-engineering/references/api-management/security-hardening-ai-gateway.md
+++ b/skills/ms-ai-engineering/references/api-management/security-hardening-ai-gateway.md
@@ -1,6 +1,6 @@
 # Security Hardening for AI Gateways in APIM
 
-**Last updated:** 2026-02
+**Last updated:** 2026-06-24
 **Status:** GA
 **Category:** API Management & AI Gateway
 
@@ -103,17 +103,17 @@ Prompt injection er den mest kritiske AI-spesifikke trusselen (OWASP LLM Top 10
     <inbound>
         <base />
         <!-- Azure AI Content Safety for prompt moderation -->
-        <llm-content-safety backend-id="content-safety-backend">
-            <text-blocklist-ids>
-                <id>prompt-injection-patterns</id>
-                <id>offensive-content-no</id>
-            </text-blocklist-ids>
+        <llm-content-safety backend-id="content-safety-backend" shield-prompt="true">
             <categories>
                 <category name="Hate" threshold="2" />
                 <category name="Violence" threshold="2" />
                 <category name="SelfHarm" threshold="2" />
                 <category name="Sexual" threshold="2" />
             </categories>
+            <blocklists>
+                <id>prompt-injection-patterns</id>
+                <id>offensive-content-no</id>
+            </blocklists>
         </llm-content-safety>
     </inbound>
 </policies>