Control physical access in layers
Physical access can expose servers, cables, and portable devices to theft or tampering. Fences channel pedestrian movement, bollards can restrict vehicle access, and lighting supports observation. Guards assess events and can respond. Fixed cameras cover specific views. Pan-tilt-zoom cameras can survey wider areas, and some systems alert or record when configured to detect motion or specified objects.
- Infrared
- Active infrared detects interruption of an emitted beam. Passive infrared detects changes in infrared energy, such as a warm body moving across its field of view.
- Microwave
- Uses radio waves to detect movement. Coverage and sensitivity depend on placement and the surrounding materials.
- Ultrasonic
- Uses high-frequency sound and can infer movement or distance from returned signals.
- Pressure
- Detects force or weight, such as someone stepping on a pressure mat.
A public waiting area provides a buffer. A staffed reception area adds credential checks and visitor handling. An access control vestibule uses interlocked doors to manage passage into a restricted area, typically allowing one door to open at a time during normal operation. Mechanical keys can be lost or copied. Electronic locks can support revocable credentials and access logs, depending on the installation.
A Faraday enclosure attenuates electromagnetic signals. Its effectiveness depends on how well it is sealed and which frequencies it is designed to block. Protected distribution systems safeguard sensitive circuits in transit through measures such as hardened or alarmed conduits. For portable equipment, use a compatible cable lock where appropriate and secure unattended devices and sensitive papers in a locked cabinet, safe, or vault suited to the risk.
NIST SP 800-53 Rev. 5 · Physical and Environmental Protection controls
Classify data by the protection it needs
Data classification assigns handling requirements to information. An organization may use labels such as public, private, sensitive, confidential, or restricted. Meanings and levels depend on its policy. Critical can also describe the operational impact of losing access or integrity. Public emergency instructions, for example, may be critical to operations and still intended for public release.
Data type helps determine which rules apply. Examples include regulated information, intellectual property, trade secrets, legal records, and financial records. Human-readable content such as reports and machine-readable content such as encoded records can both contain sensitive information. Define who may access each class and how it can be shared, retained, and disposed of.
A breach can expose customers or employees to fraud and create operational, financial, legal, and reputational consequences for the organization. Notification duties depend on the affected data, the incident, and applicable law.
Choose controls for each data state and location
- At rest
- Data stored in a database, on a device, or in a backup. Apply access restrictions and appropriate encryption. Control access to the keys.
- In transit
- Data moving between endpoints. Use protected transport and verify the endpoint before transferring sensitive information.
- In use
- Data actively processed or displayed. Limit which processes and users can see it, and mask unnecessary fields in displays or exports.
Data minimization limits what is collected and retained to what the purpose requires. Masking obscures part or all of a value in a particular view or copy. Tokenization substitutes a token for a sensitive value and limits access to the system that can recover or map the original. Access and transfer restrictions control who can use or send the data. Segmentation limits which networks or systems can reach it.
Track where data is stored, processed, backed up, and accessed. Geolocation identifies or estimates location. Data sovereignty concerns the legal authority that can apply to data across jurisdictions. Physical storage location is one factor. Applicable rules can also depend on the organization’s establishment, the people whose data is processed, and the activities involved. The GDPR’s territorial scope is an example.
NIST · Security and Privacy Controls for Information Systems and Organizations · EUR-Lex · GDPR, Articles 3 and 5
Practice case
A supplier changes its payment instructions. Accounts payable receives a convincing email claiming a regular supplier needs an overdue invoice paid today to a new bank account. A caller claiming to be the supplier’s technician also asks the front desk for access to an unattended finance workstation. That workstation can export a vendor file containing bank details.
Which persuasion cues are present? How would you verify the payment change and the visitor’s identity? Choose controls for the workstation and for the vendor file while it is stored, transferred, or displayed.
Review a possible answer
The sender uses urgency and the familiarity of a known supplier. The caller claims the authority of a technician. Pause the payment, contact the supplier using the number already on file, and follow the approval process for bank changes. Reception can verify the visit with the internal sponsor, log it, and require an escort before allowing entry. Keep the workstation behind controlled access and restrict exports to approved roles.
Store the vendor file with access controls and encryption. Use an approved protected channel if transfer is required. Mask bank details in routine displays and limit exports to the fields needed for the task. Report the messages through the organization’s incident process so related attempts can be checked.
Read and watch
- Review examples of payment redirection and how to verify changes. FBI · Business Email Compromise
- Browse the physical access, monitoring, and transmission protection controls. NIST · SP 800-53 Rev. 5
- Watch examples of recognizing and reporting suspicious messages. CISA video · How to Avoid Phishing
Recognize social engineering
Social engineering persuades someone to disclose information or take an action that weakens security. A targeted attacker may research a person or role, invent a plausible story (a pretext), and ask for a credential, payment, or access. Common persuasion cues include claimed authority, intimidation, consensus, scarcity, urgency, familiarity, and trust. A message that appears to come from a known supplier can use several cues at once.
Attackers can gather information through discarded documents or devices (dumpster diving), advanced searches for publicly indexed material (often called Google dorking), or observation of a screen or keypad (shoulder surfing). Misinformation is false or misleading information shared without intent to deceive in common security usage. Disinformation is deliberately deceptive, such as a fabricated security warning spread to steer users to an attacker’s site.
FBI IC3 · Business Email Compromise · Verisign research · Bit-level errors in DNS queries (PDF) · CISA Advisory Committee · Misinformation and disinformation (PDF)