Identify mobile devices and their attack surfaces
Smartphones run apps and use cellular or other networks. Tablets typically emphasize touch input, and wearables include watches and sensors worn on the body. Laptops and notebooks provide desktop-style computing in a portable form. Convertibles can switch between laptop and tablet use. Web-first computers emphasize browser and cloud applications, though some functions may work offline.
Cellular networks serve devices through coverage areas called cells. Wi-Fi connects devices to local wireless networks. Bluetooth connects nearby devices such as headsets or sensors, and USB can carry data as well as power. Manage each interface according to its purpose. Use trusted networks and accessories, restrict unnecessary pairing or USB data access, and keep device and accessory firmware supported.
Devices can be lost or stolen outside the organization’s perimeter. An unlocked device may expose work data or an active session. Older models may no longer receive security updates. Location services, cameras, and microphones can expose sensitive information when apps or attackers gain access, so review permissions and collect location data only as needed.
Jailbreaking on iOS and rooting on Android can weaken platform protections by bypassing restrictions or granting elevated access. Sideloading installs an app outside the usual store or managed channel and may be available without jailbreaking or rooting. Evaluate the app’s source, signature, and update path. Messages and QR codes can lead to phishing pages or unsafe downloads. Verify unexpected links and installation requests through a known channel.
NIST · Managing the Security of Mobile Devices in the Enterprise · FTC · QR code scams
Match management controls to the deployment model
- BYOD
- Employees use personally owned devices for approved work. Define which work data and apps the organization manages.
- COPE
- The organization owns and manages the device and permits personal use under policy.
- CYOD
- Employees choose from an approved list. Ownership, payment, and support terms depend on the program.
- Corporate-owned, work-only
- The organization purchases and manages the device and limits it to business use.
Require a suitable passcode, automatic screen lock, encryption, and supported operating system versions. Use a managed work profile or container where available to separate corporate apps and data from personal content. Define lost-device procedures before deployment. Locate or remotely lock the device where supported and authorized, revoke work access, and use a selective work-data wipe or a full wipe as policy and ownership permit. A device may need connectivity before it receives a remote command.
- Mobile device management (MDM) enrolls devices, applies configuration and compliance policies, and can issue supported remote actions.
- Mobile application management (MAM) distributes and manages work apps and can restrict how they exchange data.
- Mobile content management (MCM) controls access to work documents and can govern sharing or offline storage.
- Unified endpoint management (UEM) combines endpoint management capabilities across device classes. Features depend on the platform and enrollment model.
NIST SP 800-124 Rev. 2 · Mobile security across the device lifecycle
Secure embedded devices and industrial controls
Embedded systems perform specialized tasks, from reading sensors to controlling industrial equipment. Developers can assemble prototypes from readily available boards and modules. A system on a chip (SoC) integrates components such as a processor, interfaces, and controllers on one integrated circuit. A real-time operating system (RTOS) helps tasks meet timing deadlines. SoCs can also run general-purpose operating systems or firmware without an operating system.
Industrial control systems (ICS) monitor and control physical processes. A local controller might read a sensor and operate a pump or valve. Supervisory control and data acquisition (SCADA) is a type of industrial control system that gathers telemetry and supervises processes, often across geographically dispersed sites. IoT devices use network connectivity to exchange data or accept commands. Some connect through a gateway without direct public Internet exposure.
Limited memory, long service lives, weak defaults, and narrow maintenance windows can complicate security. Inventory devices and their support status, change default credentials, restrict remote access, and segment control networks. Coordinate changes with operations staff and validate updates against safety and availability requirements. NIST publishes an IoT cybersecurity capability baseline. The EU Cyber Resilience Act is an example of legislation covering products with digital elements within its scope.
NIST · Guide to Operational Technology Security · NIST · IoT cybersecurity baseline · EUR-Lex · Cyber Resilience Act
Build application security into the lifecycle
The software development lifecycle includes requirements, design, implementation, testing, deployment, and maintenance. DevSecOps, also called SecDevOps, puts security work throughout that lifecycle. Define security requirements and model threats early. During implementation, verify authorization on the server, validate input at trust boundaries, use least privilege, keep secrets out of source code, and exclude sensitive data from logs.
- Static analysis (SAST)
- Examines source code, bytecode, or other supported build artifacts without running the application. It can flag unsafe handling of untrusted input or a hard-coded secret.
- Dynamic testing (DAST)
- Exercises a running application in a test environment and observes behavior. It can reveal whether a real request bypasses an access check or exposes data.
Review findings, fix the underlying defects, and retest. Static and dynamic methods complement code review, dependency checks, and tests for business rules. Dynamic testing depends on application execution. Interpreted apps can be tested dynamically as well.
NIST · Secure Software Development Framework · OWASP Developer Guide · Static Analysis
Practice case
A lost phone at a water utility. A technician loses a personal phone enrolled in the utility’s BYOD program. Its managed work app displays pump telemetry through an API. Technicians use the app to view readings for their assigned sites. Only local industrial controllers operate the pumps. While investigating, the development team learns that the API may accept a site identifier without checking whether it belongs to the requesting technician.
How would you contain the lost-device risk? How should the mobile app reach telemetry without reaching the controllers? What would static analysis and a dynamic test look for in the API?
Review a possible answer
Revoke the app’s sessions and tokens, and request a selective work-data wipe if the enrollment and policy support it. Verify what the management service reports about delivery. Serve telemetry through a controlled gateway or a replica outside the control network, with read-only access for the mobile app and no direct path to controllers.
Review the API handler and its authorization middleware to find where site access is checked. In a test environment, sign in as a technician assigned to one site and request readings for another. Confirm that the API denies the request, then keep the scenario as a regression test.
Read and watch
- Review mobile deployment and lifecycle controls. NIST · Guidelines for Managing the Security of Mobile Devices in the Enterprise
- Review industrial control architectures and the role of segmentation. NIST · Guide to Operational Technology Security
- Watch an overview of secure development practices. NIST video · Overview of the Secure Software Development Framework