Network - 25415 - Enterprise generative AI applications are protected from prompt injection attacks through AI Gateway

[Manoj-Kesana]

Enterprise generative AI applications are protected from prompt injection attacks through AI Gateway

[Copilot]

Pull request overview

Adds a new Network/Global Secure Access assessment (25415) to verify that Prompt Shield (AI Gateway) prompt-injection protections are configured and enforced.

Changes:

• Introduces Test-Assessment-25415 PowerShell test that queries prompt policies and evaluates enforcement via Baseline Profile or CA-assigned Security Profiles.
• Adds accompanying markdown guidance content and a %TestResult% insertion point for reporting.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 2 comments.

File	Description
src/powershell/tests/Test-Assessment.25415.ps1	Implements the assessment logic, Graph data collection, enforcement evaluation, and detailed markdown report tables.
src/powershell/tests/Test-Assessment.25415.md	Provides remediation steps and embeds the %TestResult% placeholder for generated output.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

Pull request overview

Copilot reviewed 2 out of 2 changed files in this pull request and generated 3 comments.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

Baseline profiles are added as "enabled" based only on ProfileState and PassesCriteria, but the policy link state is not considered. This can mark Prompt Shield as enforced even when the prompt policy link is disabled. Include a PolicyLinkState -eq 'enabled' check when adding to $enabledBaselineProfiles (similar to Test-Assessment.25408.ps1:89-94).

Suggested change

	if ($profileLink.ProfileType -eq 'Baseline Profile' -and $profileLink.PassesCriteria -and $profileLink.ProfileState -eq 'enabled') {
	if ($profileLink.ProfileType -eq 'Baseline Profile' -and $profileLink.PassesCriteria -and $profileLink.ProfileState -eq 'enabled' -and $profileLink.PolicyLinkState -eq 'enabled') {

[Copilot]

Security profiles are treated as enforcing Prompt Shield even if the prompt policy link itself is disabled, because the elseif condition doesn't check PolicyLinkState. Add -and $profileLink.PolicyLinkState -eq 'enabled' to avoid false positives when a policy is linked but disabled.

Suggested change

	elseif ($profileLink.ProfileType -eq 'Security Profile' -and $profileLink.PassesCriteria -and $profileLink.ProfileState -eq 'enabled') {
	elseif ($profileLink.ProfileType -eq 'Security Profile' -and $profileLink.PassesCriteria -and $profileLink.ProfileState -eq 'enabled' -and $profileLink.PolicyLinkState -eq 'enabled') {

[Copilot]

$enabledSecurityProfiles only contains entries where ProfileType -eq 'Security Profile', so the computed Is Baseline column will always be "No". Removing this column (and the isBaseline calculation) would simplify the report and avoid confusion.

Suggested change

	$securityProfilesSection += "| Profile Name | State | Priority | Prompt Policy | CA Policies Assigned | Is Baseline |`n"
	$securityProfilesSection += "| :----------- | :---- | :------- | :------------ | :------------------- | :---------- |`n"
	foreach ($profile in $enabledSecurityProfiles) {
	$profilePortalLink = "https://entra.microsoft.com/#view/Microsoft_Azure_Network_Access/EditProfileMenuBlade.MenuView/~/basics/profileId/$($profile.ProfileId)"
	$policyPortalLink = "https://entra.microsoft.com/#view/Microsoft_Azure_Network_Access/EditPromptPolicyMenuBlade.MenuView/~/basics/policyId/$($profile.PromptPolicyId)"
	$profileName = Get-SafeMarkdown -Text $profile.ProfileName
	$policyName = Get-SafeMarkdown -Text $profile.PromptPolicyName
	$isBaseline = if ($profile.ProfilePriority -eq $BASELINE_PROFILE_PRIORITY) { 'Yes' } else { 'No' }
	$caCount = $profile.CAPolicyCount
	$securityProfilesSection += "| [$profileName]($profilePortalLink) | $($profile.ProfileState) | $($profile.ProfilePriority) | [$policyName]($policyPortalLink) | $caCount | $isBaseline |`n"
	$securityProfilesSection += "| Profile Name | State | Priority | Prompt Policy | CA Policies Assigned |`n"
	$securityProfilesSection += "| :----------- | :---- | :------- | :------------ | :------------------- |`n"
	foreach ($profile in $enabledSecurityProfiles) {
	$profilePortalLink = "https://entra.microsoft.com/#view/Microsoft_Azure_Network_Access/EditProfileMenuBlade.MenuView/~/basics/profileId/$($profile.ProfileId)"
	$policyPortalLink = "https://entra.microsoft.com/#view/Microsoft_Azure_Network_Access/EditPromptPolicyMenuBlade.MenuView/~/basics/policyId/$($profile.PromptPolicyId)"
	$profileName = Get-SafeMarkdown -Text $profile.ProfileName
	$policyName = Get-SafeMarkdown -Text $profile.PromptPolicyName
	$caCount = $profile.CAPolicyCount
	$securityProfilesSection += "| [$profileName]($profilePortalLink) | $($profile.ProfileState) | $($profile.ProfilePriority) | [$policyName]($policyPortalLink) | $caCount |`n"