Block AI Apps on Managed Mobile Devices with MDM

The Challenge

Why Mobile AI Apps Are a Blind Spot for Enterprise Security

Mobile devices operate outside your perimeter. When an employee opens ChatGPT on their iPhone over cellular data, that traffic never touches your firewall, proxy, or DLP engine.

The Hidden AI App Ecosystem

The threat extends far beyond obvious chatbot apps like ChatGPT, Claude, and Gemini.

AI Note-Taking Apps

Transcribe and summarize meetings, capturing confidential discussions.

AI Email Clients

Draft responses using AI, sending email content to third-party APIs.

AI Camera & Image Apps

Process images through cloud AI services, potentially exposing sensitive visuals.

Productivity Tools with Hidden AI

Quietly send document content to AI APIs for "smart" features without disclosure.

MDM: The Enforcement Layer Network Security Cannot Reach

MDM platforms provide device-level control regardless of network connection type.

Jamf Pro Apple device fleets

Microsoft Intune Cross-platform fleets

Workspace ONE Enterprise-scale deployments

Device-level enforcement, everywhere. Through configuration profiles, app restriction policies, managed browser controls, and per-app VPN tunnels, you block AI apps whether the device is on corporate Wi-Fi, home network, or cellular data. Our blocklist provides 17,410+ AI-tool domains and app bundle identifiers, classified into 18 categories and updated daily.

iOS Deployment

Block AI Apps on iOS with Restriction Profiles and Managed App Configurations

Apple's MDM framework provides two primary enforcement mechanisms for supervised iOS devices.

App Restriction Profiles

Prevent installation or launch of specific apps by bundle identifier. Delivered as XML configuration payloads pushed to enrolled devices.

Managed Browser Configs

Block access to AI-tool web domains through Safari and third-party managed browsers at the OS level.

How Bundle-ID Blocking Works

Restriction profiles use the com.apple.applicationaccess payload type. On supervised devices, you define a blocklist of app bundle identifiers — if a blocked app is already installed, it becomes hidden and non-launchable until the restriction is removed.

Commonly Blocked Bundle IDs

ChatGPT — com.openai.chatgpt
Claude — com.anthropic.claude
Gemini — com.google.Bard
Microsoft Copilot — com.microsoft.copilot
Perplexity — ai.perplexity.app.ios

Allowlist mode for high-security environments. The same payload supports a whitelist mode where only approved bundle IDs can be installed. Ideal for government agencies, defense contractors, and financial institutions where devices are locked to specific business applications.

Jamf Pro: iOS App Restriction Configuration Profile

In Jamf Pro, navigate to Devices > Configuration Profiles > New Profile > Restrictions.

1

Open the Apps Tab

Select Restrict App Usage and set mode to Do not allow some apps.

2

Add Bundle Identifiers

Enter IDs manually or upload a list from our API export.

3

Deploy via API (Optional)

Use the Jamf Pro API to push the configuration profile programmatically for automated deployment.

iOS App Restriction Profile

XML plist · Jamf / ABM / Any MDM

<!-- iOS App Restriction Profile — Block AI Applications -->
<!-- Deploy via Jamf Pro, ABM, or any MDM supporting Apple Configuration Profiles -->

<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN"
  "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
<plist version="1.0">
<dict>
  <key>PayloadContent</key>
  <array>
    <dict>
      <key>PayloadType</key>
      <string>com.apple.applicationaccess</string>
      <key>PayloadVersion</key>
      <integer>1</integer>
      <key>PayloadIdentifier</key>
      <string>com.corp.ai-blocklist.restrictions</string>
      <key>PayloadUUID</key>
      <string>A1B2C3D4-E5F6-7890-ABCD-EF1234567890</string>
      <key>blockedAppBundleIDs</key>
      <array>
        <!-- AI Chatbots -->
        <string>com.openai.chatgpt</string>
        <string>com.anthropic.claude</string>
        <string>com.google.Bard</string>
        <string>com.microsoft.copilot</string>
        <string>ai.perplexity.app.ios</string>
        <string>com.quora.app.poe</string>
        <!-- AI Image Generators -->
        <string>com.midjourney.ios</string>
        <string>com.stability.diffusion</string>
        <string>com.lensa.app</string>
        <!-- AI Writing Tools -->
        <string>com.jasper.ios</string>
        <string>ai.copy.app</string>
        <string>com.grammarly.ios</string>
        <!-- AI Code Assistants -->
        <string>com.github.copilot</string>
        <string>com.replit.app</string>
        <!-- AI Voice/Audio -->
        <string>com.elevenlabs.app</string>
        <string>com.descript.app</string>
      </array>
    </dict>
  </array>
  <key>PayloadDisplayName</key>
  <string>AI Tools Blocklist — App Restrictions</string>
  <key>PayloadIdentifier</key>
  <string>com.corp.ai-blocklist</string>
  <key>PayloadType</key>
  <string>Configuration</string>
  <key>PayloadVersion</key>
  <integer>1</integer>
</dict>
</plist>

Keeping bundle IDs current. New AI apps appear on the App Store daily, each with a unique bundle identifier — manually tracking them is unsustainable at scale. Our AI Tools Blocklist maintains a continuously updated database of AI app bundle identifiers for both iOS and Android, organized by 18 categories. Consume this data through our API and programmatically update your MDM profiles daily — newly discovered AI apps are blocked within 24 hours.

Supervised Mode Required

App restriction profiles require iOS supervised mode, set during enrollment through Apple Business Manager (ABM) or Apple School Manager (ASM).

User-initiated MDM enrollment is not supervised — those devices can only enforce managed app configurations and web content filters.

Web Content Filter Fallback

For non-supervised and BYOD devices, use the com.apple.webcontent-filter payload to block AI-tool domains at the browser level.

This filters Safari and any app using the system WebView — no supervision required.

Microsoft Intune

Configure AI App Blocking in Microsoft Intune for iOS and Android

Intune is the most widely deployed enterprise MDM, included with Microsoft 365 E3 and E5 licenses. It provides multiple enforcement mechanisms for blocking AI apps.

Mechanism	Device Type	How It Works
App Protection Policies (MAM)	BYOD + Corporate	Block managed app access if AI apps detected
Device Compliance Policies	Enrolled devices	Flag devices with AI apps as non-compliant
Device Configuration Profiles	Enrolled devices	Push app restriction payloads with blocked bundle IDs
Conditional Access	All (via Entra ID)	Gate M365 access on device compliance status

Android Enterprise

Configure a managed Google Play store controlling which apps are available. Apps not explicitly approved cannot be installed.

This allowlist approach automatically excludes every AI app — current and future — because they were never approved.

iOS via ADE

For iOS devices enrolled through Automated Device Enrollment, Intune pushes configuration profiles with blocked bundle identifiers.

Compliance policies detect AI apps and trigger Conditional Access to block Exchange, SharePoint, and Teams.

Intune App Protection Policy with Conditional Launch

Microsoft Graph API · JSON

App protection policies (APP) provide control even on unenrolled BYOD devices. Through conditional launch settings, you block access to managed corporate apps if AI apps are detected.

# Microsoft Graph API — Create Intune App Protection Policy
# Blocks access to managed apps if AI tools are detected on the device

POST https://graph.microsoft.com/v1.0/deviceAppManagement/
     managedAppPolicies

{
  "@odata.type": "#microsoft.graph.iosManagedAppProtection",
  "displayName": "Block AI Tools — iOS APP",
  "description": "Prevents corporate data access when AI apps are installed",
  "periodOfflineBeforeAccessCheck": "PT12H",
  "allowedDataStorageLocations": ["oneDriveForBusiness", "sharePoint"],
  "dataBackupBlocked": true,
  "managedBrowserToOpenLinksRequired": true,
  "appActionIfUnableToAuthenticateUser": "block",
  "thirdPartyKeyboardsBlocked": true,
  "apps": [
    {"mobileAppIdentifier": {
      "@odata.type": "#microsoft.graph.iosMobileAppIdentifier",
      "bundleId": "com.microsoft.Office.Outlook"}},
    {"mobileAppIdentifier": {
      "@odata.type": "#microsoft.graph.iosMobileAppIdentifier",
      "bundleId": "com.microsoft.teams"}}
  ],
  "exemptedAppProtocols": [],
  "conditionaLaunchRules": [
    {
      "settingName": "BlockedApps",
      "appBundleIds": [
        "com.openai.chatgpt",
        "com.anthropic.claude",
        "com.google.Bard",
        "com.microsoft.copilot",
        "ai.perplexity.app.ios"
      ],
      "action": "block",
      "warningMessage": "AI applications must be removed to access corporate data."
    }
  ]
}

How This Policy Works

Targets Managed Apps Only

Blocks access to Outlook, Teams, and OneDrive if AI app bundle IDs are detected. Users must uninstall AI apps to regain access.

Works on BYOD Without Enrollment

No full device enrollment or supervision required. It manages corporate apps and data, not the entire device.

Near-Instantaneous Detection

The conditional launch check runs when the user opens a managed app, ensuring immediate enforcement.

Android equivalent. Use androidManagedAppProtection with package names: com.openai.chatgpt, com.anthropic.claude, com.google.android.apps.bard, com.microsoft.copilot, ai.perplexity.app.android. Our API provides both iOS bundle IDs and Android package names in a single data source.

Compliance Policies

Flag devices running AI apps as non-compliant. Non-compliant devices are blocked from corporate resources through Conditional Access in Microsoft Entra ID.

Managed Browser

Configure Microsoft Edge with a URL blocklist including all AI-tool domains. Users are blocked and redirected to a policy explanation page.

Conditional Access

Pair compliance policies with Entra Conditional Access. Devices with AI apps lose access to Exchange, SharePoint, and Teams until apps are removed.

Mobile AI Threat Landscape

The App Stores Ship AI Faster Than You Can Review It

Every one of these apps carries its own bundle ID, its own backend domains, and its own data-handling terms — and the count grows every day.

2,000+AI Apps on iOS App StoreChatbots, cameras, keyboards, and more

3,500+AI Apps on Google PlayEach with a unique package name

67%Use AI Outside Corp NetworkCellular and home Wi-Fi bypass your perimeter

DailyNew AI Apps AppearingTracked and classified in our feed

VMware Workspace ONE

VMware Workspace ONE: App Compliance and Smart Group Enforcement

Workspace ONE (formerly AirWatch) uses a compliance engine that continuously monitors enrolled devices for policy violations. The engine runs on a configurable schedule — as frequent as every 15 minutes.

Setup in Three Steps

1

Create an Application Group

Navigate to Apps & Books > Application Groups > Add Group. Name it AI Tools — Blocked and set the type to Blocklist. Add app identifiers manually or import a CSV from our API export.

2

Create a Compliance Policy

Under Devices > Compliance Policies, set the rule to Application List > Contains Blocklisted Apps and select your application group.

3

Configure the Escalation Chain

Define a multi-step response for progressive enforcement.

Escalation Timeline

Immediately Push notification explaining the policy violation. User has 24 hours to remove the app.

After 24 Hours Block access to corporate email and Wi-Fi profiles.

After 48 Hours Enterprise wipe — removes all corporate data and profiles while preserving personal data.

Smart Groups for Granular Policy Assignment

Smart Groups let you target AI blocking policies to specific user segments. Filter by organizational group, AD user group, device platform, ownership type, and more.

Department-Based Smart Groups

Target policies at specific Active Directory OUs. Each Smart Group can reference a different application group.

Block all AI tools for Finance, Legal, HR
Block only AI writing tools for Marketing
Allow everything for Engineering and R&D

Ownership-Based Smart Groups

Differentiate policies by device ownership model.

Corporate: app restriction profiles preventing AI app installation entirely
BYOD: compliance policies that detect AI apps and restrict corporate resource access

Managed Browsers

Managed Browser Configurations: Blocking AI Web Access on Mobile

Blocking AI native apps is only half the battle — every AI tool also has a web interface, and a user who finds ChatGPT blocked as an app will open Safari and navigate to chat.openai.com. Managed browser configurations close this gap by extending your AI domain blocklist to the browser layer on managed devices.

Platform-Specific Browser Controls

iOS Content Filter

Apple provides the com.apple.webcontent-filter payload with built-in filter and plugin modes.

Add AI-tool domains to the blacklisted URLs array. iOS blocks access in Safari and any app using WKWebView or system URL loading.

Android Managed Browsers

Chrome supports the URLBlocklist managed configuration with JSON URL patterns.

Microsoft Edge uses Intune app configuration. VMware Browser uses the Workspace ONE SDK for content filtering rules.

Chrome Managed Configuration for Android Enterprise

JSON · Intune / WS1

Deploy this JSON payload as a managed configuration through Intune, Workspace ONE, or any MDM supporting Android Enterprise. Our API exports domain lists compatible with Chrome's URL pattern syntax for automated daily updates.

// Chrome Managed Configuration — Android Enterprise
// Deploy via Intune App Configuration or WS1 Managed Config

{
  "URLBlocklist": [
    // AI Chatbots & Assistants
    "chat.openai.com",
    "openai.com",
    "claude.ai",
    "anthropic.com",
    "gemini.google.com",
    "bard.google.com",
    "copilot.microsoft.com",
    "perplexity.ai",
    "poe.com",
    "you.com",
    // AI Image Generation
    "midjourney.com",
    "stability.ai",
    "dreamstudio.ai",
    "dall-e.com",
    "leonardo.ai",
    // AI Writing & Content
    "jasper.ai",
    "copy.ai",
    "writesonic.com",
    "rytr.me",
    // AI Code Assistants
    "github.com/features/copilot",
    "replit.com",
    "cursor.sh",
    "codeium.com",
    // AI Voice & Audio
    "elevenlabs.io",
    "descript.com",
    "murf.ai"
  ],
  "URLBlocklistExceptions": [
    // Allowlisted internal AI tools
    "ai.internal.yourcompany.com"
  ],
  "BrowserSigninAllowed": true,
  "SafeBrowsingProtectionLevel": 2,
  "IncognitoModeAvailability": 1
}

Critical: Prevent Bypass

Disable Incognito Mode — IncognitoModeAvailability: 1 — without this, users can open incognito tabs that bypass managed configurations entirely.
Enable Enhanced Safe Browsing — SafeBrowsingProtectionLevel: 2 — adds an extra defense layer against AI-tool domains classified as potentially unsafe.

Safari on Managed iOS

URL blocklists are delivered through the web content filter profile, not browser-specific managed configs.

This filters all web traffic on the device — Safari and any app using the system WebView. Domain entries must be exact matches or subdomains.

Advantage vs. Chrome

iOS content filtering operates at the OS level, catching traffic from any app — not just the browser.

Trade-off: no wildcard pattern support like Chrome's URLBlocklist key allows.

Network-Level Controls

Per-App VPN Policies: Route AI App Traffic Through Your Security Stack

Per-app VPN routes traffic from specific apps through a VPN tunnel to your corporate network. Your existing firewalls, proxies, and DLP engines inspect and control it.

Network-Level Blocking

AI app traffic passes through the VPN tunnel, hits your firewall, matches AI-tool domains in the blocklist, and is dropped. Users see a connection timeout.

Shadow AI Visibility

See which AI apps users attempt to use — even apps not on your blocklist. Critical intelligence for policy development.

How It Works

1

Configure a Per-App VPN Profile

Apply to all apps or a targeted set, routing traffic through a VPN tunnel to your corporate network.

2

Firewall Matches AI Domains

Your SWG or firewall has the AI Tools Blocklist loaded as a URL filtering category and drops matching connections.

3

No Device-Level Management Needed

Ideal for BYOD — you don't need to know what AI apps are installed. Any app reaching an AI-tool domain is blocked at the network level.

Split Tunnel vs. Full Tunnel

Full tunnel: routes all traffic through corporate network. Complete visibility, higher bandwidth cost.
Split tunnel: routes only AI-tool traffic. Our API provides the complete list of domains and IP ranges for split-tunnel configuration.

Shadow AI Discovery

Start with monitor-only mode. Analyze firewall logs for connections to AI-tool domains before enforcing blocks.
Some AI tools may be embedded in business workflows and need managed exceptions rather than blanket blocks.

Deployment Strategy

BYOD vs. Corporate-Owned: Choosing the Right AI Blocking Strategy

Your blocking strategy depends on the device ownership model. Each gives you different levels of control.

Capability	Corporate-Owned	BYOD
Supervised / Device Owner mode	Yes	No
App installation blocking	Yes	No
Device-level app scanning	Yes	Requires consent
MAM / App protection policies	Yes	Yes
Conditional Access	Yes	Yes
Enterprise wipe	Full + Selective	Selective only

Corporate-Owned Devices

App restriction profiles with blocked bundle IDs
Supervised mode (iOS) / Device Owner mode (Android)
Managed App Store with allowlist-only installation
Full-device web content filtering
Per-app or full-tunnel VPN for network-level blocking
Compliance policies with enterprise wipe escalation
Managed browser as the only permitted browser
Kiosk or single-app mode for frontline devices

BYOD Devices

App protection policies (MAM) on managed apps only
Conditional access: block corp data if AI apps detected
Managed browser for corporate web access only
Work profile (Android) for data separation
Copy/paste restrictions between managed and personal apps
No device-level app scanning without consent
Data containerization rather than device control
User education and acceptable use policy enforcement

Recommended: Dual-Strategy Deployment

The most effective approach uses both strategies simultaneously. Corporate devices get the full blocking stack; BYOD devices get MAM with conditional access.

One feed, every format. Both device types consume the same AI Tools Blocklist data. Our API exports in multiple formats: iOS bundle IDs, Android package names, domain lists, and JSON feeds for programmatic MDM API integrations.

The Managed Browser Gateway Model

A proven hybrid approach used by large enterprises. Corporate resources open only in the managed browser (Edge, Chrome Enterprise, or VMware Browser), which has the AI-tool URL blocklist applied.

Users can still use personal browsers for personal browsing — including AI tools. But the MAM policy blocks cut-copy-paste between managed and unmanaged apps.
Result: AI tools are technically accessible on the device, but corporate data cannot flow into them. Data containment without device control.

Automation

Automating MDM Policy Updates with the AI Tools Blocklist API

New AI tools launch on the App Store and Google Play every week. Manually updating MDM restriction profiles is unsustainable at scale. Our API keeps your policies current automatically.

The Automation Pipeline

1

Scheduled Script Fetches Latest Data

Run as a cron job, Azure Function, AWS Lambda, or CI/CD pipeline. Calls our API daily for the latest AI app identifiers and domains.

2

Transform to MDM Format

XML configuration profiles for Jamf, JSON payloads for Intune's Graph API, CSV imports for Workspace ONE.

3

Push Updated Policy via MDM API

The entire pipeline runs unattended. Your MDM policies are always current.

Jamf Pro API

Fetch AI app bundle IDs, generate mobileconfig XML, and upload via the Jamf Pro API. Assign profiles to Smart Groups for targeted deployment.

Microsoft Graph API

Update Intune app protection and compliance policies. Our JSON output maps directly to Intune's policy schema — minimal transformation needed.

WS1 UEM API

Push new app identifiers through the UEM API to update application groups. Triggers compliance re-evaluation across all enrolled devices.

For MDM Vendors

Our API supports bulk data access with real-time webhook notifications. When a new AI app is discovered, a webhook fires with full metadata — name, bundle ID, package name, category, risk score, and domain associations.

Your platform can consume these webhooks and update its built-in AI app database without waiting for a daily batch job. Several MDM vendors already use this integration model to power native AI app blocking features.

Ready to Block AI Apps on Your Managed Devices?

Download the free sample to test with your MDM platform today. Or tell us your deployment details and we will send a ready-to-import app restriction list within 24 hours.

Download Free Sample View Plans

Request Your MDM AI App Blocklist

Tell us your MDM platform, device fleet size, and OS mix. We will deliver a formatted app restriction list within 24 hours.

Keep Exploring

Block AI Apps on Managed Mobile Devices