CertifyDeck
All cheat sheets
SY0-801 · CHEAT SHEET

CompTIA Security+

Key concepts grouped by official exam weight.

Download as PDF (free, no sign-up)

1. General Security Concepts

16–16%

1.1 Explain security concepts and controls.

  • Defense in depth layers people, technology and operations barriers so that what one layer misses another catches. The CIA triad protects confidentiality, integrity and availability; AAA separates authentication, authorization and accounting (records of what was done). Non-repudiation prevents denying an action, least privilege limits rights and Zero Trust decides each access without implicit trust.
  • Control categories describe who or what applies a control. Technical controls are executed by hardware, software or firmware mechanisms (firewalls, account lockout); managerial controls focus on managing risk and the security program (risk assessments, policies); operational controls are executed by people (training, help desk procedures) and physical controls protect facilities and equipment.
  • Control types describe when and how a control acts: preventive stops the incident (USB blocking, IPS), deterrent discourages the attacker (warning signs), detective finds what happened (IDS, log review), corrective restores the situation (restoring backups), directive tells people what to do (policies) and compensating gives comparable protection when the normal control cannot be applied.

1.2 Given a scenario, demonstrate the impact of change management processes on security.

  • A change is requested, recorded, analyzed for security impact, tested and approved by the change advisory board (CAB), which receives the test plan and results. The request names an owner, the stakeholders to notify, a maintenance window and a backout plan. Detected changes are reconciled by checking for an approved request and whether they occurred in the scheduled window.
  • A change has technical consequences to plan for: new software must be added to the allow list under deny-all, permit-by-exception; dependencies on other systems are part of impact analysis; legacy applications may fail under hardened configurations; and the implementer is limited to approved activities, with notice given when restarts or downtime will occur.
  • A change is not finished when it is deployed: network diagrams, technical designs, the baseline configuration, inventory, security plan and the procedures used by the help desk and users must be updated. This applies to pre-approved standard changes too. The previous baseline and approval records are kept unchanged for traceability, audits and possible rollback.
  • Version control keeps code, configuration as code and baseline configurations in a repository that records every change and the account that made it. Keeping previous versions makes it possible to return to an approved configuration and helps audits and incident response. Because history keeps everything, a committed secret stays exposed even if deleted later: it must be revoked and rotated.

1.3 Explain the importance of using appropriate cryptographic solutions.

  • In a PKI each identity has a key pair: the public key can be shared and is used to encrypt data for the owner and verify the owner's signatures; the private key is never published and is used to sign and decrypt. Key escrow retains the private key tied to an encryption certificate so data can be recovered if the user loses it or leaves the organization.
  • A certificate authority issues and revokes certificates, and its root is the starting point of trust. A CRL is a signed, periodic list of revoked certificates, while OCSP checks one certificate's status with more timely information. A CSR carries the public key signed with the private key. A self-signed certificate depends on secure distribution, and a wildcard covers one left-most label only.
  • Symmetric encryption (AES) uses the same key to encrypt and decrypt and handles large volumes; asymmetric encryption (RSA, ECC) establishes keys and signs in hybrid schemes. Disk, partition or volume encryption protects against theft of a powered-off device but not on a running system; database or record encryption protects specific values. TLS 1.3 allows only key exchanges with forward secrecy.
  • A digital signature is generated with the signer's private key and verified with the public key; it proves origin and data integrity and provides non-repudiation because only the signer holds the private key. An HMAC uses a shared key and gives no non-repudiation. Code signing with certificates from an established CA lets operating systems check who published an executable and whether it was altered.
  • A salt is a unique random value combined with each password before hashing it with a slow password hashing scheme. It makes identical passwords produce different hashes and defeats precomputed rainbow tables. It is stored with the hash and need not be secret; the secret value kept elsewhere is a pepper. Slowness comes from the cost factor, not the salt.

2. Threats, Vulnerabilities, and Attacks

24–24%

2.1 Explain characteristics of threats and vulnerabilities.

  • Threat intelligence is threat information that has been aggregated, analyzed and enriched to give context for decisions; a raw feed of indicators is not intelligence on its own. In a risk assessment, likelihood estimates whether a threat can exploit a vulnerability, and impact measures the magnitude of harm if it does. Internal sources (sensors, logs) are complemented by external feeds and sharing communities.
  • CVE gives a common identifier to each publicly known vulnerability and CVSS measures its severity: Base metrics are intrinsic, while Threat and Environmental metrics are tuned by the consumer to its own case. The qualitative bands are Low 0.1-3.9, Medium 4.0-6.9, High 7.0-8.9 and Critical 9.0-10.0. For prioritization, CISA advises fixing first what appears in the KEV catalog, which has proven active exploitation.

2.2 Describe common threat actors and motivations.

  • CISA defines an insider as anyone who has or had authorized access to or knowledge of the organization's resources, including contractors and vendors with access. Insider threat can be intentional (malicious), negligent or accidental, such as sending a document to a competitor by mistake. External actors such as hacktivists, unskilled attackers or state groups operate without that legitimate access.
  • Motivation explains what the actor wants: financial gain, espionage, notoriety, revenge, extortion or an ideological or political cause, among others. CISA notes that many intentional insiders want to get even after a dismissal or a lack of recognition, and that some hacktivist groups mainly seek notoriety and exaggerate their attacks on social media to attract attention.
  • Actors are distinguished by whether they are internal or external, by their resources and funding and by their level of sophistication. NIST SP 800-30 assesses adversaries by capability (expertise, resources, opportunities), intent and targeting. CISA also describes collusive threats, in which an insider works with an external actor, often cybercriminals who recruit the insider to commit fraud.

2.3 Describe threat vectors and sources.

  • NIST defines an APT as a sophisticated, well-resourced adversary that uses several vectors (cyber, physical, deception) and pursues its objectives over long periods, adapting to defenses. CISA links it to nation-state actors with aims such as espionage, data theft or disruption, and warns that one APT can carry several names because each organization assigns its own.
  • Message-based vectors arrive through email, SMS, RCS, instant messaging or collaboration tools. Smishing uses SMS with urgency so the victim taps a link or installs malware. With a compromised internal account, an attacker can run internal spearphishing over corporate chat, abusing trust in known senders to steal credentials on fake login pages.
  • Image-based vectors hide the lure inside an image: quishing uses QR codes that lead to phishing sites and pivot from the computer to the phone, evading URL inspection. Fake CAPTCHAs (ClickFix) ask the user to paste and run a command, which sidesteps email filtering and browser sandboxing; application control helps stop them.
  • Malicious attachments (documents with macros, RTF, PDF, archives) depend on the user opening them. A macro virus is embedded in a document and activates when it is handled, so disabling macros is advisable. At the mail gateway, blocking unnecessary attachment types by default and scanning and removing malicious attachments before delivery are recommended.
  • The browser is a vector through its extensions, JavaScript, saved passwords and session cookies and tokens. A stolen session cookie lets an attacker use a web application as the user and bypass some MFA, because the session is already authenticated. Malicious extensions mimic legitimate names; they are mitigated with extension allow lists and by installing only from trusted sources.

2.4 Explain types of vulnerabilities and attack surfaces.

  • An unsupported product no longer receives vendor updates, so every new flaw stays unfixed; CISA warns that unsupported edge devices are targeted by exploitation campaigns and should be phased out. Until they are replaced, NIST SP 800-40r4 treats them as unpatchable: isolate them, for example with micro-segmentation, and periodically reassess the mitigations.
  • Patch management identifies, prioritizes, acquires, installs and verifies patches. A patch may need a service or platform restart to take effect, so it must be verified as truly applied and monitored to stay installed. If a flaw is actively exploited and no patch exists yet, NIST recommends a temporary emergency mitigation, later replaced by the fix.
  • Obsolete systems use technology that no longer meets current requirements: legacy devices without encryption, logging or passwords and without resources to add security, or deprecated protocols such as TLS 1.0 and 1.1, which the IETF retired because they lack current algorithms. Until they are replaced, they must be protected with compensating controls in the surrounding network.
  • An unmanaged system is an asset connected to the network that sits outside inventory, patching, configuration and monitoring processes, so its vulnerabilities pile up unseen. CIS Control 1 calls for actively inventorying, tracking and correcting all enterprise assets (end-user, network, IoT, servers and cloud) to find unauthorized or unmanaged ones and remove or remediate them.
  • Every open port and listening service widens the attack surface, especially remote administration services exposed to the internet. NIST SP 800-123 recommends removing, rather than merely disabling, unnecessary services, and MITRE ATT&CK advises routing remote access through VPNs or other managed concentrators with MFA instead of exposing it directly.

2.5 Given a scenario, analyze indicators of malicious activity.

  • Each type of malware leaves different indicators: ransomware encrypts and renames files and demands a ransom, a worm spreads across the network on its own, a logic bomb waits for a condition to trigger, a rootkit hides processes and files to keep privileged access, spyware secretly gathers information and fileless malware runs in memory without leaving files on disk.
  • Physical attacks seek access or information without going through the network. In tailgating, an unauthorized person walks in behind an authorized one; in shoulder surfing, someone watches what the victim types or sees on screen. They are mitigated with awareness so staff let nobody through and require badges, physical access controls and obscuring authentication feedback on screen.
  • Network attacks have their own indicators: a DDoS saturates bandwidth from many sources, an on-path attack relies on ARP poisoning or downgrading HTTPS to HTTP, and DNS cache poisoning makes a resolver accept forged answers and repeat them to its clients. RFC 5452 recommends randomizing IDs and source ports to make forgery harder.
  • Social engineering is classified mainly by channel: phishing through email or fake websites, smishing through SMS with links and vishing by voice, often with spoofed caller ID. CISA and MITRE ATT&CK point to impersonating a trusted source and creating urgency as common traits, and warn that vishing is combined with MFA requests to steal codes.
  • An indicator of compromise is an artifact or observable suggesting an attack is imminent, under way or has already succeeded: malicious hashes, IPs or domains, odd processes, altered timestamps, cleared logs, excessive resource consumption, account lockouts or impossible travel. MITRE ATT&CK describes how attackers tamper with timestamps and logs to hide.

2.6 Summarize threats and vulnerabilities associated with artificial intelligence (AI) usage.

  • The AI threats in this group affect model integrity: poisoning manipulates training or fine-tuning data to introduce bias or backdoors, and prompt injection uses direct input or external content to alter behavior and cause data loss. Mitigations include verifying data provenance, separating untrusted content and requiring human approval for risky actions.
  • Evasion modifies inputs at deployment time so a model misclassifies them without touching training, and hallucination produces fabricated content that looks correct, such as nonexistent software packages an attacker can register. Membership inference threatens privacy, and jailbreaking is a prompt injection that makes the model ignore its safety rules.
  • When LLM output is passed unvalidated to other components, classic risks appear: run in a shell it leads to remote code execution, and displayed unencoded in a browser it enables XSS and session token theft. Treat the model as an untrusted user, validate its output and encode it for the context where it is used.

3. Security Architecture

19–19%

3.1 Compare and contrast security implications of different architecture models.

  • Cloud deployment models differ by who uses them: a private cloud serves one organization, a community cloud a group with shared requirements, a public cloud is open to everyone and a hybrid cloud combines several infrastructures. An air-gapped network has no physical connection to others and exchanges data only manually, unlike logical segmentation with VLANs or firewalls.
  • When comparing architectures, consider responsibility: in IaaS the customer manages operating systems, storage and applications, while in PaaS the provider handles the operating system. Resilience is the ability to keep operating under adverse conditions, even degraded, and recover within mission time frames; availability is timely and reliable access to information.
  • Business considerations for an architecture include data sovereignty, because data is subject to the laws of where it is processed and the GDPR restricts transfers to third countries, and scalability, which the cloud provides through rapid elasticity to provision and release capacity with demand.

3.2 Given a scenario, manage the security architecture to best protect the infrastructure.

  • Device placement defines security zones: public servers belong in a DMZ or screened subnet, which gives restricted outside access while shielding the internal network. The attack surface is the set of points where an attacker can enter or extract data, and it shrinks by disabling unneeded services, ports and software.
  • In Zero Trust no user or asset is trusted for being on the network: each access is granted per session under dynamic policy that considers identity and device posture. The policy engine decides, the policy administrator executes the decision and the policy enforcement point enables, monitors and terminates connections to the resource.
  • Secure communication includes deciding which traffic uses the VPN: split tunneling protects only internal traffic and prevents inspecting the rest, so it should be disabled on untrusted networks. Out-of-band management uses dedicated paths to administer network devices, separates that traffic from user traffic and allows fixes without an attacker on the production network seeing them.
  • Identity management applies least privilege to service accounts too: a gMSA has a password generated and rotated automatically by the directory, which hinders attacks such as Kerberoasting. Privilege creep appears when someone changes roles and keeps previous permissions; it is prevented with periodic access reviews and by removing unneeded permissions.
  • Failure modes define what a system does when it fails: fail-secure (closed) keeps the secure state even if service stops, fail-safe prioritizes preventing harm to people or property even at the cost of security, and fail-soft stops only non-essential functions. The choice depends on whether the requirement prioritizes information security or human safety.

3.3 Summarize concepts and strategies used to protect data.

  • Structured data follows a predefined schema of rows and fields, such as a database table, while unstructured data, such as free text, images, audio and video, lacks explicit structure relating data to individuals. That makes unstructured data harder to locate, classify and protect with field-based tools.
  • Data has three states: at rest when stored on a device or medium, in transit when moving between systems or sites, and in use when being viewed, modified or processed. Each state needs different controls: disk encryption at rest and protocols such as TLS 1.3 or IPsec in transit; TLS 1.0 and 1.1 are deprecated.
  • Classification assigns each piece of data a level based on the harm its disclosure or loss would cause. Public information is what the organization makes available to everyone; sensitive information could harm privacy or operations without being officially classified, and levels such as secret or top secret apply only to information designated as classified for national security.
  • Several methods protect data depending on how it will be used: masking hides part of a value and does not preserve its analytic utility, pseudonymization replaces identifiers with pseudonyms and keeps a controlled way to re-link them, and anonymization removes the association with the person. Encryption and tokenization replace the whole value.
  • Data protection roles split responsibilities: the owner decides classification and sets the controls for the information, the custodian applies them day to day, the controller determines purposes and means, and the processor processes data on its behalf. Under the GDPR, a processor cannot engage a subprocessor without the controller's prior written authorization.

3.4 Explain the importance of resilience and recovery in security architecture.

  • Alternate sites are chosen by recovery time and cost: a cold site has only space, power and environmental controls; a warm site adds equipment without current software or data; a hot site is fully operational, and a mirrored site replicates in real time. The site should also be in an area not exposed to the same hazards as the primary site.
  • Platform diversity, such as using different vendors, hardware or hypervisors, keeps a single shared vulnerability from causing a common mode failure across all components. NIST warns that it can be more apparent than real if products share components, and that it can increase the attack surface and management costs, so it must be analyzed.
  • Redundancy removes single points of failure: load balancing spreads traffic, clustering lets another node take over, and autoscaling adds or removes capacity with demand. For power, dual supplies cover the failure of one unit, a UPS gives short runtime for a graceful shutdown, and a generator covers long outages.
  • Resilient storage combines RAID, replication and snapshots. Synchronous replication copies data in real time to the DR site and avoids losing committed writes; asynchronous replication adds a delay tied to the RPO. Snapshots are point-in-time copies that depend on the source data, so they do not replace a separate copy.
  • A backup policy defines scope (full, incremental or differential), a frequency tied to the RPO, and retention. Immutable or offline backups resist ransomware that deletes or encrypts accessible copies, and only restoration testing and integrity checks with hashes prove that backups can be recovered.

4. Security Operations

27–27%

4.1 Given a scenario, apply mitigating controls, techniques, and solutions to secure the environment.

  • Segmentation divides the network into isolated zones, using VLANs, subnets or separate physical networks, and controls traffic between them with firewalls, routers or cloud rules. It reduces the attack surface, slows lateral movement and protects critical assets such as payment systems or IoT and OT devices.
  • Access controls decide who or what can use a resource and with which permissions. ACLs enumerate the authorized entities on routers, firewalls or file systems, and least privilege limits permissions to what is needed, for example by making administrators use non-privileged accounts for email and browsing.
  • Hardening reduces the attack surface: patching, disabling unneeded services, ports and protocols, removing unused software and changing default settings. Consensus-based prescriptive guides such as the CIS Benchmarks provide a baseline for configuring operating systems and applications consistently.
  • Sandboxing runs untrusted code in an isolated, controlled environment, such as a virtual machine, a container or the browser sandbox, with permissions cut to the essentials. It contains threats so they cannot affect the rest of the system, and detonation chambers use it to analyze attachments and URLs before delivery.
  • Deception technology plants decoys with no legitimate use, so any interaction exposes an attacker. Honeypots and honeynets are systems or networks that attract adversaries and let defenders study them; honeyfiles, honeytokens and canary accounts are fake files, credentials or accounts that alert when used. They must be isolated from production.

4.2 Explain the security implications of proper hardware, software, and data asset management.

  • An asset's life cycle starts before purchase: planning and acquisition set security requirements in the contract, request the software's SBOM and buy from manufacturers, authorized distributors or authorized resellers to avoid counterfeits. Buying on the secondary market calls for a prior risk assessment.
  • When an asset is enrolled it is assigned an owner and its serial number, location and baseline configuration are recorded; during operation, agents and network monitoring track changes and detect unenrolled devices. At retirement its data is sanitized with clear, purge or destroy according to sensitivity and the media's destination, and it is removed from the inventory.

4.3 Given a scenario, perform tasks associated with vulnerability management.

  • Vulnerabilities are identified with complementary methods: host and network vulnerability scanning, static source code analysis (SAST), which flags flaws such as SQL injection with file and line, and CSPM, which continuously monitors cloud service configuration and permissions. SAST cannot see configuration issues that are not in the code.
  • The CVSS base score measures a vulnerability's intrinsic severity; prioritization should add threat metrics, such as active exploitation, and environmental metrics, such as system criticality and existing controls. CISA's KEV catalog flags what is already exploited, and in a pentest report findings that give control of critical systems come first.
  • A vulnerability's risk can be accepted, mitigated, transferred or avoided. Patching or upgrading is the only response that removes the flaw without losing functionality; when no patch exists, compensating controls such as network segmentation mitigate it. Exceptions to the maintenance plan are documented, monitored and reviewed regularly.
  • Verifying remediation means confirming that the patch is installed and has taken effect (sometimes a restart is needed), that added controls work and that retired assets are gone. For a test finding, the same original test case is repeated; a result from another tool or profile is not comparable.
  • Internal vulnerability reports are tailored to each audience and include an action plan. Externally, a vulnerability disclosure policy says where to report, which systems and tests are authorized and promises no legal action against good-faith research; a bug bounty adds payment for valid findings.

4.4 Explain security alerting and monitoring concepts and tools.

  • Monitoring covers systems, applications and infrastructure. Operating system logs record system events and audit records (logons, file access, account changes) and help investigate a specific host. Application logs show what each user did inside the application, which over encrypted connections only the application itself can record.
  • Monitoring activities include aggregating and normalizing logs centrally, alerting, scanning, archiving and reporting. In archiving, retention is routine while preservation keeps logs that would be discarded because they matter to an investigation. Alert tuning reduces false positives without creating false negatives and must be reviewed periodically.
  • Monitoring tools include configuration benchmarks such as CIS's, the SIEM that centralizes and correlates logs, antivirus, DLP, vulnerability scanners, orchestration and packet analyzers. Agentless collection avoids installing software on each host but sends more unfiltered data and may need credentials; agents filter at the source but must be maintained.
  • NetFlow exports flow records (IP addresses, packet and byte counts) without packet payloads. SNMP monitors and manages devices, and SNMPv3 adds authentication and encryption. Syslog carries event messages to a central collector. SCAP standardizes the format of checklists and vulnerability data (CVE, CPE, CVSS, OVAL, XCCDF) so they can be assessed automatically.
  • Automated alerts allow a response within minutes to events of interest, such as privileges being assigned to an account, instead of waiting for a manual review. They rely on automated filtering of known log entries and on thresholds over a baseline of normal activity, for example a sudden rise in outbound email volume.

4.5 Given a scenario, apply concepts related to identity and access management.

  • Provisioning requires an approved request and permissions based on a valid authorization for the user's duties. Deprovisioning must be tied to HR processes: account managers are notified of terminations and transfers, unneeded accounts are disabled and shared account credentials are changed when someone leaves the group.
  • Permissions follow least privilege: only the access needed for current tasks. They must be reviewed periodically, removing those left over after a transfer to prevent privilege creep. Separation of duties splits incompatible tasks, such as creating vendors and approving payments, between roles to prevent fraud without collusion.
  • Identity proofing collects, validates and verifies information to establish confidence in a person's claimed identity: identity resolution, validation of evidence and attributes, verification that the applicant owns the evidence, and enrollment. IAL1 and IAL2 allow remote processes; IAL3 requires an on-site session with a trained agent and a biometric.
  • Federation conveys identity and authentication information across systems in different domains: the identity provider (IdP) authenticates the user and sends an assertion to the relying party (RP) under a trust agreement. The RP validates the signature with the expected IdP's key, checks it is the intended audience and rejects reused assertions.
  • SSO lets one account and its authenticators be used across several applications, usually through a federation protocol. SAML exchanges XML assertions between partners; OAuth 2.0 grants a third-party application limited access through tokens, without sharing the password; LDAP provides directory access, and its password simple bind must be protected with TLS (StartTLS).

4.6 Given a scenario, apply automation and orchestration solutions to secure operations.

  • Automation and scripting apply to user provisioning (joiners, movers and leavers driven by HR), resource provisioning, desired state management (keeping the approved configuration), anomaly detection and ticket management, for example creating and assigning tickets automatically when a certain type of alert fires.
  • Before automating, weigh guardrails (human approval for high-impact actions), the logic and complexity of workflows, cost, process risks such as false positives that stop services, and deployment: automated changes are tested, validated and documented outside production and still follow change control.
  • AI in security operations includes agents that take actions through tools, chatbots that assist analysts, predictive analysis and AI-augmented baselines that learn normal behavior. Agents need minimal permissions and human approval for high-impact actions, and baselines depend on clean training data.
  • Automation aims at concrete outcomes: efficiency and time saving, enforcing approved configuration baselines, productivity and continuous improvements, reduced downtime and increased proactivity, for example freeing analysts for threat hunting. It is not meant to remove human review from decisions.
  • In CI/CD, continuous integration automates build and test, continuous delivery promotes to staging and leaves the production push manual, and continuous deployment also automates production. The pipeline widens the attack surface, so it needs secrets kept out of code in a dedicated manager, least privilege, change review and artifact integrity.

4.7 Summarize concepts associated with incident response activities.

  • Preparation happens before an incident: policies with defined roles and authorities (who may disconnect systems), procedures written as playbooks, staff training and plan testing. Tabletop exercises are discussion-based and touch no systems, while functional exercises and simulations have the team act in a simulated environment.
  • Identification detects and confirms incidents from alerts and continuous monitoring, internal advisories, external advisories from suppliers or third parties with their IoCs, and proactive threat hunting driven by threat intelligence. Analyzed activity is compared with incident criteria to decide whether to declare an incident.
  • Investigation applies digital forensics while preserving data integrity. Chain of custody records who handled each piece of evidence, when and why; preservation keeps data that would normally be deleted, and e-discovery requires locating and producing the information requested in litigation without altering it, suspending its scheduled deletion.
  • Containment stops an incident from spreading: quarantining malware, isolating hosts on remediation networks or halting infected containers. It should be coordinated, using out-of-band channels if the attacker may be watching communications, and powering hosts down is best avoided unless they cannot be disconnected, because memory evidence is lost.
  • In an extortion case, paying the attackers guarantees neither file recovery nor that stolen data stays unpublished, since double extortion is common. Before deciding, consult law enforcement, which may know of decryptors already available, and preserve evidence.

4.8 Given a scenario, use data, artifacts and sources to support a security investigation.

  • Each log type answers different questions: authentication logs show the origin, user, result and time of each attempt; audit logs show account changes and file access; application logs show requests and responses; network logs show connections; physical access logs show facility entries; and metadata gives details such as a file's author or modification time.
  • Investigation data sources are chosen by the question: NetFlow or IPFIX records summarize who talked to whom and how many bytes over long periods; packet captures keep content but only briefly; vulnerability scans show weaknesses; surveillance footage records physical presence; and automated reports and dashboards summarize what security tools collect.
  • File and log integrity lets them be trusted as evidence. Hashes reveal any change, file integrity checking software compares hashes of critical files, and logs are protected by sending them to a separate system, limiting access to audit tools and using cryptographic protection such as signed hashes.
  • A system image preserves evidence: a memory dump captures volatile data lost at power-down, such as memory-resident malware; a bit-level copy duplicates the whole medium, including deleted files and slack space, and is verified with hashes; and a snapshot records the state of a virtual machine or cloud volume at a point in time.
  • An investigation involves non-technical parties: the legal department reviews plans and contracts and advises when an incident may lead to legal action or requires decisions such as delaying containment; human resources is involved when an employee is suspected; auditors help estimate the economic impact and physical security helps with access and securing evidence.

5. Security Program Management and Oversight

14–14%

5.1 Explain the importance of governance, risk, and compliance artifacts.

  • Guidelines advise rather than mandate: benchmarks (such as the CIS Benchmarks) are prescriptive, consensus-based configuration recommendations for hardening systems; security advisories report new vulnerabilities with affected products and mitigations; and a reference architecture is a vendor-neutral conceptual model used as the basis for designing specific architectures.
  • Standards set mandatory, measurable requirements. A control baseline is the minimum set of controls for a low-, moderate- or high-impact system; a password standard aligned with NIST SP 800-63B-4 favors length and blocklist checks and forbids composition rules, periodic expiration without compromise and hints. RFCs are IETF standards documents.
  • Procedures (SOPs, runbooks, playbooks) explain step by step how to perform a technical process and derive from the policy and plan. They should be documented for the most common incidents and for urgent processes, tested or exercised periodically to verify they are still accurate, and used to train new staff.
  • The business continuity plan (BCP) sustains business processes, such as payroll, during and after a disruption and may be scoped to priority functions. The disaster recovery plan (DRP) is system-focused: it moves system operations to an alternate site when the primary facility is inaccessible, after which each system is recovered with its own contingency plan.
  • Policies capture the organization's high-level rules. A vulnerability disclosure policy tells outside researchers the scope, the unauthorized methods, how to report and the commitment not to take legal action against good-faith research; a data disposal policy maps each classification level to a minimum sanitization method (clear, purge or destroy).

5.2 Explain the impact of risk management processes on the security of the organization.

  • Risk identification starts by inventorying the assets that support the mission and valuing them beyond replacement cost, because without that information impact cannot be quantified. Threats, vulnerabilities and consequences are then analyzed. Each registered risk has an owner who is accountable for it and chooses the response, and who may delegate monitoring to a risk manager.
  • FIPS 199 categorization assigns each objective (confidentiality, integrity, availability) a potential impact of LOW, MODERATE or HIGH, and a system takes the highest value of its information types for each objective. For scoring, a semi-quantitative scale (0-100 in bins) translates easily into qualitative levels and allows risks to be compared, although each bin must be well defined.
  • Risk analysis combines likelihood and impact. Likelihood for adversarial threats rests on intent, capability and targeting; for non-adversarial threats, on historical and empirical data, taking deployed controls into account. Qualitative analysis uses nonnumerical levels that ease communication; quantitative analysis uses numbers with their own meaning and best supports cost-benefit analysis.
  • The risk register is the central repository of current risks, both accepted and under mitigation, with description, likelihood, impact, response, owner and status. It is a communication tool: each area's registers are normalized and aggregated into an enterprise register and prioritized into a profile for leadership. It must be reviewed continually, including accepted risks.
  • There are four risk responses: accept it when it is within tolerance; avoid it by eliminating or changing the activity or technology that creates it; mitigate it with controls that reduce likelihood or impact; and transfer or share it, for example through insurance, which shifts liability but reduces neither likelihood nor consequences.

5.3 Explain the assessment and management processes associated with third-party risk.

  • Vendor selection starts with due diligence: requests for information (RFIs) and questionnaires to screen candidates and learn their risk posture. A statement of work is then released with a request for proposal (RFP) or quote (RFQ), and responses are evaluated against security criteria, ensuring no evaluator has a conflict of interest.
  • Each agreement has a role: the SLA commits the provider to performance levels, responsibilities and reporting; the SOW details the contractor's work and deliverables; the MOU records a mutual understanding without payment or reliance, while the MOA documents support with or without reimbursement; and the NDA defines information that must not be disclosed.
  • Vendor monitoring continues after the contract: a supplier inventory or registry with the products, the systems that use them and their criticality; periodic revalidation of security requirements through certifications, site visits, third-party assessments or self-attestation, with rigor proportional to criticality; and contract clauses on audit and information exchange.
  • Third-party management has constraints: vendor lock-in (proprietary environments, cost of migrating), the applicable jurisdiction and regulations, lack of visibility into the provider's controls, reliance on a single source or region, and available resources. They are offset with generalized interfaces, exit plans, independent third-party audits and continuous monitoring.
  • Rules of engagement (ROE) are agreed before a security test and authorize the team to perform activities that are normally prohibited. They cover scope, schedule, personnel and contacts, target systems and the exclude list, allowed activities, data handling, reporting and signatures, plus the criteria for halting testing if an incident or real intruder appears.

5.4 Summarize elements of effective security compliance.

  • Compliance training complements general awareness with role-based training for people who handle sensitive information: it is delivered before access to systems is granted and refreshed periodically. It covers data handling, protection of personal information and reporting of incidents or policy violations.
  • Compliance monitoring relies on documented evidence. Acknowledgements confirm that each person has read and accepts rules such as the rules of behavior, before access is granted and again when they change. Attestations formally state that requirements are met, such as the PCI DSS AOC, which summarizes the results of a ROC or SAQ.
  • Non-compliance has several consequences: regulator sanctions, such as GDPR administrative fines or a limitation or ban on processing; legal liability, such as compensation that affected people claim in court; financial costs; contractual penalties; and reputational damage from the loss of trust of customers and partners.
  • The GDPR grants data subjects rights: rectification of inaccurate data, erasure or the right to be forgotten, restriction of processing (for example, while contested accuracy is verified) and objection, as well as withdrawing consent. The controller determines the purposes and means of processing; the processor processes the data on the controller's behalf and instructions.
  • Legal compliance shapes data retention: retention schedules must meet legal and regulatory needs; when litigation is expected, a legal hold requires keeping unaltered records that would normally be deleted; and the GDPR right to erasure gives way when a law requires certain data to be kept. Obtaining third-party records, such as an ISP's, usually requires a court order.

5.5 Explain concepts associated with audit and assessment activities.

  • Audits gather evidence by examining documents and configurations, interviewing people and testing controls, often over a representative sample. As reference sources, MITRE ATT&CK documents adversary tactics and techniques, the Cyber Kill Chain orders adversary objectives into phases, and the Diamond Model helps cluster related intrusions.
  • Assessment scope sets which systems are covered, how large the target is and how often each one is assessed. Frequency usually comes from compliance requirements and risk, and priority combines the impact rating with the date of the last assessment; when resources fall short, high-priority systems come first and the scope is reduced.
  • An audit can be internal, run by the organization itself (compliance reviews, audit committee work, self-assessments), or external, run by a second or third party such as a regulator or an independent auditor. A gap analysis compares the current state with the desired one, while benchmarking compares it with other organizations or industry practice.
  • In a penetration test, assessors try to defeat security features under agreed rules; depending on what they know beforehand, the environment is known, partially known or unknown. Passive reconnaissance observes traffic without sending probes, while active reconnaissance sends packets, is faster and can trigger IDS alerts.
  • Audits rely on frameworks and standards of different reach: industry-based, such as PCI DSS for environments holding payment card data; international, such as ISO/IEC 27001, which sets the requirements for a certifiable information security management system; and region-specific, such as the European Union GDPR for personal data.

5.6 Given a scenario, apply security awareness concepts to improve organizational security.

  • Security training combines several types: initial, at onboarding and before access is granted; ongoing, with recurring sessions and updates on recent threats; targeted, for groups or roles with higher risk, such as those failing simulations most often; and corrective, delivered promptly after an audit finding or an incident.
  • Training can be delivered through a learning management system (LMS), which distributes content and records registration and completion; self-service portals with on-demand material; one-to-one sessions such as mentoring; or one-to-many formats such as classroom sessions or live webinars. Blending several mechanisms helps hold the audience's attention.
  • Awareness effectiveness is measured with metrics that go beyond attendance or completion, such as the percentage of users who report a simulated phishing email. Managerial reports present these trends to leadership, and personnel behavior risk scoring helps identify the people with higher human risk so they can receive non-punitive support.
  • Common awareness topics include social engineering and phishing, business email compromise (BEC), where payment requests are verified through a known contact, password management, remote work and BYOD, and removable media: a found USB drive is handed to security or IT without plugging it in.