Skip to main content
Be Bitwise

Compliance & Threat Response

Mobile Enterprise & MDM Security · Mobile Advanced Security track

10 min read

In this module

From Policy to Enforcement

Enterprise mobile security holds up exactly as long as you can spot non-compliance and respond before data leaves the building. This lesson walks through the compliance frameworks that define "secure" on mobile, the machinery that enforces compliance round the clock, and the incident response playbook for the bad days.

NIST SP 800-124: The Mobile Security Baseline

NIST Special Publication 800-124 (Guidelines for Managing the Security of Mobile Devices in the Enterprise) is the canonical reference for enterprise mobile security. The 2023 revision pushes a layered approach:

  • Centralised management: All mobile devices accessing corporate data must be enrolled in an MDM/EMM platform.
  • Device-level controls: Full-disk encryption, strong authentication (PIN/biometric with minimum complexity), automatic lock after inactivity, and remote wipe capability.
  • Network-level controls: VPN for corporate traffic, certificate-based WiFi authentication, and network access control.
  • Application management: Enterprise app store with vetted applications, managed app configuration, and app-level data protection.
  • Incident response: Documented procedures for lost, stolen, and compromised devices.

CIS Benchmarks

The Centre for Internet Security (CIS) publishes configuration benchmarks for iOS and Android, and they go granular:

  • Minimum passcode length and complexity.
  • Auto-lock timeout (recommended: 2 minutes or less).
  • Whether to allow biometric authentication.
  • Camera and screenshot restrictions in managed contexts.
  • Bluetooth and NFC restrictions.
  • Maximum failed passcode attempts before wipe.

CIS Benchmarks are blunter than NIST. Where NIST recommends, CIS dictates exact values. Most organisations end up using NIST as the policy framework and CIS Benchmarks as the day-to-day implementation guide.

Device Compliance Policies

MDM platforms ride on device compliance policies — the floor that a device has to clear before getting near corporate resources.

Common Compliance Rules

RulePurposeExample
Minimum OS versionEnsure security patches are appliedCurrent major release or the one before
Encryption requiredProtect data at restMust be enabled
Jailbreak/root detectionDetect compromised OS integrityMust not be jailbroken/rooted
Passcode requiredPrevent unauthorised accessMinimum 6 digits, alphanumeric
Maximum passcode ageForce periodic credential rotation90 days (debated)
MDM profile presentVerify device is managedMust be enrolled
MTD app installedEnsure threat detection is activeMust be installed and running
Prohibited apps absentBlock known-risky applicationsNo sideloaded APKs

Compliance Evaluation

Compliance isn't a one-shot check at enrolment. It runs continuously:

  1. The MDM agent on the device periodically reports device state (OS version, encryption status, installed apps, jailbreak indicators).
  2. The MDM server evaluates the reported state against the compliance policy.
  3. If compliant, the device retains access to corporate resources.
  4. If non-compliant, the MDM triggers an escalation chain.
Continuous Compliance

Compliance is not a point-in-time check. A device that was compliant yesterday may be non-compliant today because the user disabled encryption, installed a prohibited app, or the OS version fell below the minimum after a new security update was released. Continuous compliance evaluation catches these changes and can revoke access within minutes, limiting the window of exposure.

Non-Compliance Escalation

A device drops out of compliance, the response follows a ladder:

  1. Notification: Alert the user with remediation instructions ("Your OS version is outdated. Update within 48 hours to maintain access").
  2. Access restriction: If not remediated within the grace period, revoke access to corporate resources via Conditional Access.
  3. Selective wipe: If the device remains non-compliant or the risk is severe (jailbreak detected), wipe corporate data.
  4. Full wipe: Reserved for lost/stolen devices or severe compromise on corporate-owned devices.

Remote Wipe

Remote wipe is the panic button. Lost, stolen, popped — MDM platforms back two flavours:

Selective Wipe (Enterprise Wipe)

Selective wipe removes only enterprise-managed data:

  • Android: Deletes the Work Profile and all data within it. Personal apps and data are untouched.
  • iOS: Removes managed apps and their data, corporate email accounts, WiFi/VPN profiles, and certificates. Personal apps and data remain.

Selective wipe is the right move on BYOD, where the organisation has no business deleting someone's personal data.

Full Wipe (Factory Reset)

Full wipe drops the device back to factory state. Personal data, corporate data, all of it gone. The right call for lost or stolen corporate hardware, or for personally-owned devices where the user signed up for full wipe in the MDM enrolment terms.

The decision has one more input that's easy to forget: evidence. A wipe destroys exactly what an investigation needs, so a device that may have been compromised gets cut off from corporate access first and wiped only once the evidence question is settled. On a personal device, any collection beyond the work profile needs the owner's consent.

Lost Device Procedures

"My phone's gone" should never trigger improvisation. Run the playbook:

  1. User reports device missing: Initiate the lost device procedure.
  2. Remote lock: Immediately lock the device with a custom lock screen message ("This device belongs to [Company]. Please call [number]").
  3. Locate: Use MDM's location tracking to determine the device's last known position.
  4. Grace period: Allow a defined period (e.g., 24-48 hours) for the device to be recovered.
  5. Wipe decision: If not recovered, execute selective wipe (BYOD) or full wipe (corporate). If there is evidence of active compromise, wipe immediately without a grace period.
  6. Credential rotation: Revoke the device's certificates, rotate any passwords that may have been stored on the device, and invalidate active SSO sessions.

Walk through it with a real timeline. Decide what you'd do at each timestamp, and why, before opening the answer.

  • Mon 09:00: The CFO's company-owned iPhone fails compliance: it's two OS updates behind. A 72-hour grace period starts.
  • Wed 18:10: The phone last checks in with the MDM, from a city-centre location.
  • Wed 18:40: The CFO reports it left in a taxi. It's still within the grace period, so it still has access.
  • Thu 09:00: No sign of it. The MDM shows no check-in since Wednesday.
Show one reasonable answer

At 18:40, lock it remotely with a contact message and mark it non-compliant by hand, which cuts corporate access through Conditional Access straight away. Don't wait for the grace period, which was there for patching, not for a missing phone. Revoke the CFO's SSO sessions and refresh tokens at the same time, since the phone's apps hold live ones. Check the last location, then give the recovery attempt a short window. By Thu 09:00 nothing has checked in, so queue the full wipe (it's corporate-owned, so that's allowed) to run on the next check-in, revoke the device's identity certificate, and rotate anything the phone could hold that a wipe doesn't fix: the CFO's password, any shared Wi-Fi or VPN secrets, and app-specific passwords. Write down each timestamp and decision as you go; the compliance failure on Monday will be the first question in the post-incident review.

App Vetting

Before any app reaches a managed device, somebody should look at it.

Enterprise App Store

An enterprise app store is a curated catalogue of pre-approved applications. Users install from there instead of the public store. The MDM can lock the Work Profile down to enterprise-catalogue apps only.

App Reputation Services

Lookout, Zimperium, Pradeo and friends will tear apart an APK or IPA looking for:

  • Known malware signatures.
  • Risky permission combinations.
  • Data collection practices (which third-party SDKs are embedded, what data they collect).
  • Network behaviour (which domains the app contacts, whether it uses certificate pinning).
  • Code quality issues (use of deprecated APIs, insecure crypto implementations).

These services hook into MDM platforms so flagged or banned apps get blocked without anyone clicking a button.

OWASP MASTG

The OWASP Mobile Application Security Testing Guide (MASTG) is as close as the industry gets to an agreed-upon methodology for mobile app security review. For enterprise app vetting, MASTG covers test cases for:

  • Data storage security (keychain usage, no plaintext secrets).
  • Network security (TLS configuration, certificate pinning).
  • Authentication and session management.
  • Code quality and platform security.

Incident Response for Mobile

Mobile IR has its own rhythm. Several things change once the device in question fits in a pocket.

Evidence Preservation

Mobile devices are volatile. Remote wipes can fire, apps can auto-purge messages, batteries die. Preserve evidence accordingly:

  • Network isolation: Cut the device off from the network so no remote wipe or attacker command reaches it. A Faraday bag is the most reliable option. Airplane Mode is weaker: it can be toggled from the lock screen, a managed device's policy may switch radios back on, and Wi-Fi or Bluetooth can be re-enabled while it's set. A phone in a Faraday bag searches hard for signal and drains its battery fast, so connect a shielded power source or acquire quickly.
  • Keep it powered, and unlocked if it already is: A reboot flushes any volatile implant from memory and puts a modern phone back into its before-first-unlock state, where most data stays encrypted. Whether losing the implant is acceptable depends on the investigation goal, so make that call deliberately. Mind the clock on iOS too: a phone left locked for about 72 hours restarts on its own.
  • Forensic acquisition: Create a forensic image using appropriate tools before any other analysis. iOS backups (encrypted, to capture keychain data) provide baseline forensic data. Android ADB backups help less than they used to: apps targeting Android 12 or later leave their data out unless they're debuggable.

Forensic Acquisition Methods

  • Logical acquisition: An encrypted iOS backup, MDM-facilitated collection, or (on older Android apps) ADB backup. Captures app data, messages, call logs, but not deleted content.
  • Filesystem acquisition: Requires jailbreak (iOS) or root (Android), or an exploit-based extraction tool. Captures the full filesystem, and deleted records often survive inside SQLite databases. It doesn't cover unallocated space.
  • Physical acquisition: Direct chip-off or JTAG extraction. The most invasive option, and on modern phones with file-based encryption and hardware-bound keys the raw image is mostly ciphertext: the keys live in the SoC's secure hardware, not on the flash chip, so the image is only useful if the device itself can later be made to unlock it. It still matters for older and low-end devices.

Containment

Mobile containment plays out differently from network containment:

  • Revoke access tokens: Immediately invalidate all SSO sessions and OAuth tokens associated with the device.
  • Revoke certificates: Revoke the device's identity certificate to prevent it from authenticating to corporate resources.
  • Quarantine in MDM: Mark the device as non-compliant, triggering Conditional Access to block all corporate resource access.
  • Selective wipe: If corporate data exposure is confirmed, execute selective wipe to remove corporate content.
Key Concept
Mobile Incident Response

Mobile incident response requires preserving volatile evidence (Faraday bag, keep the device powered), revoking all associated credentials and tokens immediately, performing forensic acquisition before any remediation, and executing selective or full wipe only after evidence is preserved. The decision to wipe must balance data protection (removing corporate data from a compromised device) against forensic value (evidence needed for investigation). Organisations should document this decision process in advance, not during the stress of an active incident.