ISC Engineer cheat sheet

CIA triad: confidentiality (disclosure prevention), integrity (unauthorized modification prevention), availability (timely reliable access).

Non-repudiation requires proof of origin binding sender to message, typically via digital signature, not symmetric MAC alone.

Defense in depth layers controls so a single failed control does not yield full compromise; each layer assumes prior layers fail.

Least privilege: subject receives only the minimum rights needed for the task, for the minimum duration.

Threat vs. vulnerability vs. risk: threat exploits, vulnerability enables, risk = likelihood x impact of exploitation.

Identification asserts an identity claim; authentication verifies it; authorization grants permissions; accountability ties actions to identities via audit.

Fail-safe defaults: a control denying access on failure (fail-closed) is preferred for confidentiality; fail-open favors availability.

Separation of duties splits a critical task across two or more subjects so no single subject can complete it alone.

Classification labels (public, internal, confidential, restricted) drive handling, retention, and encryption requirements.

DAC: resource owner sets access; MAC: system-enforced labels (e.g., clearance vs. classification); RBAC: permissions assigned to roles, users to roles.

ABAC evaluates attributes (subject, resource, action, environment) via policy; it is more granular than RBAC but harder to audit.

Authentication factors: something you know, have, are; MFA requires factors from at least two distinct categories.

SSO trades credential count for blast radius: compromise of the identity provider session affects all federated relying parties.

SAML is XML-based browser SSO; OAuth 2.0 delegates authorization (not authentication); OIDC adds an identity layer on OAuth 2.0.

Kerberos: symmetric ticket-based; TGT from KDC, service tickets from TGS; time synchronization is required to limit replay.

Access review/certification: periodic recertification of entitlements; orphaned or excessive accounts are removed at review.

Password storage: salted, slow hash (bcrypt/scrypt/Argon2); never reversible encryption; rate-limit and lock out on repeated failures.

Zero trust: no implicit trust from network location; every request is authenticated and authorized per session.

DMZ/screened subnet hosts internet-facing services; internal firewalls restrict DMZ-to-internal traffic to required flows.

Stateful firewall tracks connection state; stateless ACLs must permit return traffic explicitly.

NGFW adds application identification and user awareness; WAF inspects HTTP payloads for injection/XSS signatures.

Network segmentation limits lateral movement; VLANs are logical isolation, not a security boundary against a compromised host on the same segment.

IPsec: AH provides integrity/authentication only; ESP provides confidentiality plus integrity; IKE negotiates SAs (main/aggressive, quick mode).

TLS 1.3 removes static RSA key exchange and legacy ciphers; forward secrecy comes from ephemeral (EC)DHE.

IDS detects and alerts (passive); IPS blocks inline; signature-based vs. anomaly-based detection trade false negatives against false positives.

802.1X port-based access control: supplicant, authenticator (switch), authentication server (RADIUS) using EAP.

Microsegmentation applies per-workload policy, often host-based, to restrict east-west traffic inside a data center.

Secure SDLC: threat modeling at design, static analysis (SAST) on source, dynamic analysis (DAST) on running apps, dependency scanning for components.

Threat modeling (e.g., STRIDE) enumerates spoofing, tampering, repudiation, information disclosure, DoS, elevation of privilege per trust boundary.

Input validation is allow-list based at trust boundaries; output encoding prevents injection in the target interpreter context.

Hardening baseline: remove unused services, change defaults, patch, restrict administrative interfaces, enable logging.

Sandboxing/least privilege for processes limits impact of exploited code; memory-safe languages reduce buffer-overflow classes.

Configuration management drift detection compares running state to a known-good baseline; IaC scanning catches misconfig before deploy.

Secure boot and code signing establish a chain of trust from firmware to OS to application.

Vulnerability remediation prioritization uses exploitability plus exposure plus asset criticality, not CVSS alone.

Symmetric (AES, ChaCha20) is fast for bulk data; asymmetric (RSA, ECC) is used for key exchange and signatures, not bulk encryption.

AES modes: GCM/CCM provide authenticated encryption (confidentiality + integrity); ECB leaks patterns and must not be used for data.

Hashes (SHA-256/384/512) give one-way fixed-length digests; MD5 and SHA-1 are broken for collision resistance.

Digital signature = hash of message encrypted/signed with sender's private key; verified with sender's public key; provides integrity, authentication, non-repudiation.

HMAC uses a shared secret with a hash for message authentication; it does not provide non-repudiation.

Key lifecycle: generation, distribution, storage, rotation, revocation, destruction; private keys never leave a protected store (HSM/KMS).

PKI trust chain: root CA signs intermediate CA, which signs end-entity cert; validation checks chain, validity dates, revocation (CRL/OCSP), and name.

Forward secrecy: compromise of long-term private key does not expose past session keys when ephemeral key exchange is used.

Entropy matters: keys must come from a CSPRNG; weak randomness breaks otherwise strong algorithms.

Vulnerability scan identifies known weaknesses by signature; penetration test chains findings to demonstrate real impact.

Scan types: authenticated (credentialed) finds more, unauthenticated mimics external attacker; both need scoping and safe-rate limits.

False positive vs. false negative: tuning detection trades one against the other; measure with precision and recall.

Log sources for monitoring: auth logs, EDR, network flow, DNS, cloud audit trails; centralized SIEM correlates and alerts.

Detection engineering: map rules to ATT&CK techniques to find coverage gaps; alert on behavior, not just indicators.

Risk scoring of findings uses CVSS base plus environmental modifiers and threat intelligence (EPSS/KEV) for prioritization.

Sampling and evidence: test results must be reproducible with scope, method, timestamp, and tool versions recorded.

IR lifecycle: preparation, detection/analysis, containment, eradication, recovery, lessons learned; documentation spans all phases.

Containment options: isolate host, block IP, disable account, revoke tokens; choose based on blast radius vs. business impact.

Order of volatility: capture RAM, network state, then disk, then backups/logs; preserve chain of custody with hashes.

RTO is maximum tolerable downtime; RPO is maximum tolerable data loss; both drive backup frequency and DR architecture.

BIA identifies critical processes and dependencies; recovery strategies are sized to RTO/RPO.

DR site types: hot (near-zero RTO), warm (partial), cold (long RTO, restore from backup).

Tabletop exercises validate plans and communications; after-action reviews produce tracked corrective actions.

Backups must be tested for restore; immutable/offline copies defend against ransomware encryption.

Risk = likelihood x impact; treatment options: mitigate, transfer, avoid, accept (documented with owner and expiry).

Qualitative vs. quantitative risk: quantitative uses monetary values (ALE = SLE x ARO; SLE = asset value x exposure factor).

Control types: preventive, detective, corrective, compensating; categories: administrative, technical, physical.

Frameworks: ISO 27001 (ISMS requirements), NIST CSF (identify/protect/detect/respond/recover), SOC 2 (trust criteria), PCI DSS (card data).

Policy hierarchy: policy (intent) -> standard (mandatory rules) -> procedure (steps) -> guideline (recommended).

Third-party risk: due diligence, contractual security requirements, right-to-audit, and periodic reassessment.

Compliance evidence must be current and traceable to the control it satisfies; gaps are tracked as findings with owners and dates.

Data privacy obligations (GDPR/CCPA) require lawful basis, data minimization, retention limits, and breach notification timelines.

CIA triad: confidentiality (disclosure prevention), integrity (unauthorized modification prevention), availability (timely reliable access).

Non-repudiation requires proof of origin binding sender to message, typically via digital signature, not symmetric MAC alone.

Defense in depth layers controls so a single failed control does not yield full compromise; each layer assumes prior layers fail.

Least privilege: subject receives only the minimum rights needed for the task, for the minimum duration.

Threat vs. vulnerability vs. risk: threat exploits, vulnerability enables, risk = likelihood x impact of exploitation.

Identification asserts an identity claim; authentication verifies it; authorization grants permissions; accountability ties actions to identities via audit.

Fail-safe defaults: a control denying access on failure (fail-closed) is preferred for confidentiality; fail-open favors availability.

Separation of duties splits a critical task across two or more subjects so no single subject can complete it alone.

Classification labels (public, internal, confidential, restricted) drive handling, retention, and encryption requirements.

CIA triad: confidentiality (disclosure prevention), integrity (unauthorized modification prevention), availability (timely reliable access).

Non-repudiation requires proof of origin binding sender to message, typically via digital signature, not symmetric MAC alone.

Defense in depth layers controls so a single failed control does not yield full compromise; each layer assumes prior layers fail.

Least privilege: subject receives only the minimum rights needed for the task, for the minimum duration.

Threat vs. vulnerability vs. risk: threat exploits, vulnerability enables, risk = likelihood x impact of exploitation.

Identification asserts an identity claim; authentication verifies it; authorization grants permissions; accountability ties actions to identities via audit.

Fail-safe defaults: a control denying access on failure (fail-closed) is preferred for confidentiality; fail-open favors availability.

Separation of duties splits a critical task across two or more subjects so no single subject can complete it alone.

Classification labels (public, internal, confidential, restricted) drive handling, retention, and encryption requirements.

CIA triad: confidentiality (disclosure prevention), integrity (unauthorized modification prevention), availability (timely reliable access).

Non-repudiation requires proof of origin binding sender to message, typically via digital signature, not symmetric MAC alone.

Defense in depth layers controls so a single failed control does not yield full compromise; each layer assumes prior layers fail.

Least privilege: subject receives only the minimum rights needed for the task, for the minimum duration.

Threat vs. vulnerability vs. risk: threat exploits, vulnerability enables, risk = likelihood x impact of exploitation.

Identification asserts an identity claim; authentication verifies it; authorization grants permissions; accountability ties actions to identities via audit.

Fail-safe defaults: a control denying access on failure (fail-closed) is preferred for confidentiality; fail-open favors availability.

Separation of duties splits a critical task across two or more subjects so no single subject can complete it alone.

Classification labels (public, internal, confidential, restricted) drive handling, retention, and encryption requirements.

CIA triad: confidentiality (disclosure prevention), integrity (unauthorized modification prevention), availability (timely reliable access).

Non-repudiation requires proof of origin binding sender to message, typically via digital signature, not symmetric MAC alone.

Defense in depth layers controls so a single failed control does not yield full compromise; each layer assumes prior layers fail.

Least privilege: subject receives only the minimum rights needed for the task, for the minimum duration.

Threat vs. vulnerability vs. risk: threat exploits, vulnerability enables, risk = likelihood x impact of exploitation.

Identification asserts an identity claim; authentication verifies it; authorization grants permissions; accountability ties actions to identities via audit.

Fail-safe defaults: a control denying access on failure (fail-closed) is preferred for confidentiality; fail-open favors availability.

Separation of duties splits a critical task across two or more subjects so no single subject can complete it alone.

Classification labels (public, internal, confidential, restricted) drive handling, retention, and encryption requirements.

CIA triad: confidentiality (disclosure prevention), integrity (unauthorized modification prevention), availability (timely reliable access).

Non-repudiation requires proof of origin binding sender to message, typically via digital signature, not symmetric MAC alone.

Defense in depth layers controls so a single failed control does not yield full compromise; each layer assumes prior layers fail.

Least privilege: subject receives only the minimum rights needed for the task, for the minimum duration.

Threat vs. vulnerability vs. risk: threat exploits, vulnerability enables, risk = likelihood x impact of exploitation.

Identification asserts an identity claim; authentication verifies it; authorization grants permissions; accountability ties actions to identities via audit.

Fail-safe defaults: a control denying access on failure (fail-closed) is preferred for confidentiality; fail-open favors availability.

Separation of duties splits a critical task across two or more subjects so no single subject can complete it alone.

Classification labels (public, internal, confidential, restricted) drive handling, retention, and encryption requirements.

CIA triad: confidentiality (disclosure prevention), integrity (unauthorized modification prevention), availability (timely reliable access).

Non-repudiation requires proof of origin binding sender to message, typically via digital signature, not symmetric MAC alone.

Defense in depth layers controls so a single failed control does not yield full compromise; each layer assumes prior layers fail.

Least privilege: subject receives only the minimum rights needed for the task, for the minimum duration.

Threat vs. vulnerability vs. risk: threat exploits, vulnerability enables, risk = likelihood x impact of exploitation.

Identification asserts an identity claim; authentication verifies it; authorization grants permissions; accountability ties actions to identities via audit.

Fail-safe defaults: a control denying access on failure (fail-closed) is preferred for confidentiality; fail-open favors availability.

Separation of duties splits a critical task across two or more subjects so no single subject can complete it alone.

Classification labels (public, internal, confidential, restricted) drive handling, retention, and encryption requirements.

CIA triad: confidentiality (disclosure prevention), integrity (unauthorized modification prevention), availability (timely reliable access).

Non-repudiation requires proof of origin binding sender to message, typically via digital signature, not symmetric MAC alone.

Defense in depth layers controls so a single failed control does not yield full compromise; each layer assumes prior layers fail.

Least privilege: subject receives only the minimum rights needed for the task, for the minimum duration.

Threat vs. vulnerability vs. risk: threat exploits, vulnerability enables, risk = likelihood x impact of exploitation.

Identification asserts an identity claim; authentication verifies it; authorization grants permissions; accountability ties actions to identities via audit.

Fail-safe defaults: a control denying access on failure (fail-closed) is preferred for confidentiality; fail-open favors availability.

Separation of duties splits a critical task across two or more subjects so no single subject can complete it alone.

Classification labels (public, internal, confidential, restricted) drive handling, retention, and encryption requirements.

CIA triad: confidentiality (disclosure prevention), integrity (unauthorized modification prevention), availability (timely reliable access).

Non-repudiation requires proof of origin binding sender to message, typically via digital signature, not symmetric MAC alone.

Defense in depth layers controls so a single failed control does not yield full compromise; each layer assumes prior layers fail.

Least privilege: subject receives only the minimum rights needed for the task, for the minimum duration.

Threat vs. vulnerability vs. risk: threat exploits, vulnerability enables, risk = likelihood x impact of exploitation.

Identification asserts an identity claim; authentication verifies it; authorization grants permissions; accountability ties actions to identities via audit.

Fail-safe defaults: a control denying access on failure (fail-closed) is preferred for confidentiality; fail-open favors availability.

Separation of duties splits a critical task across two or more subjects so no single subject can complete it alone.

Classification labels (public, internal, confidential, restricted) drive handling, retention, and encryption requirements.

CIA triad: confidentiality (disclosure prevention), integrity (unauthorized modification prevention), availability (timely reliable access).

Non-repudiation requires proof of origin binding sender to message, typically via digital signature, not symmetric MAC alone.

Defense in depth layers controls so a single failed control does not yield full compromise; each layer assumes prior layers fail.

Least privilege: subject receives only the minimum rights needed for the task, for the minimum duration.

Threat vs. vulnerability vs. risk: threat exploits, vulnerability enables, risk = likelihood x impact of exploitation.

Identification asserts an identity claim; authentication verifies it; authorization grants permissions; accountability ties actions to identities via audit.

Fail-safe defaults: a control denying access on failure (fail-closed) is preferred for confidentiality; fail-open favors availability.

Separation of duties splits a critical task across two or more subjects so no single subject can complete it alone.

Classification labels (public, internal, confidential, restricted) drive handling, retention, and encryption requirements.

CIA triad: confidentiality (disclosure prevention), integrity (unauthorized modification prevention), availability (timely reliable access).

Non-repudiation requires proof of origin binding sender to message, typically via digital signature, not symmetric MAC alone.

Defense in depth layers controls so a single failed control does not yield full compromise; each layer assumes prior layers fail.

Least privilege: subject receives only the minimum rights needed for the task, for the minimum duration.

Threat vs. vulnerability vs. risk: threat exploits, vulnerability enables, risk = likelihood x impact of exploitation.

Identification asserts an identity claim; authentication verifies it; authorization grants permissions; accountability ties actions to identities via audit.

Fail-safe defaults: a control denying access on failure (fail-closed) is preferred for confidentiality; fail-open favors availability.

Separation of duties splits a critical task across two or more subjects so no single subject can complete it alone.

Classification labels (public, internal, confidential, restricted) drive handling, retention, and encryption requirements.

CIA triad: confidentiality (disclosure prevention), integrity (unauthorized modification prevention), availability (timely reliable access).

Non-repudiation requires proof of origin binding sender to message, typically via digital signature, not symmetric MAC alone.

Defense in depth layers controls so a single failed control does not yield full compromise; each layer assumes prior layers fail.

Least privilege: subject receives only the minimum rights needed for the task, for the minimum duration.

Threat vs. vulnerability vs. risk: threat exploits, vulnerability enables, risk = likelihood x impact of exploitation.

Identification asserts an identity claim; authentication verifies it; authorization grants permissions; accountability ties actions to identities via audit.

Fail-safe defaults: a control denying access on failure (fail-closed) is preferred for confidentiality; fail-open favors availability.

Separation of duties splits a critical task across two or more subjects so no single subject can complete it alone.

Classification labels (public, internal, confidential, restricted) drive handling, retention, and encryption requirements.

CIA triad: confidentiality (disclosure prevention), integrity (unauthorized modification prevention), availability (timely reliable access).

Non-repudiation requires proof of origin binding sender to message, typically via digital signature, not symmetric MAC alone.

Defense in depth layers controls so a single failed control does not yield full compromise; each layer assumes prior layers fail.

Least privilege: subject receives only the minimum rights needed for the task, for the minimum duration.

Threat vs. vulnerability vs. risk: threat exploits, vulnerability enables, risk = likelihood x impact of exploitation.

Identification asserts an identity claim; authentication verifies it; authorization grants permissions; accountability ties actions to identities via audit.

Fail-safe defaults: a control denying access on failure (fail-closed) is preferred for confidentiality; fail-open favors availability.

Separation of duties splits a critical task across two or more subjects so no single subject can complete it alone.

Classification labels (public, internal, confidential, restricted) drive handling, retention, and encryption requirements.

CIA triad: confidentiality (disclosure prevention), integrity (unauthorized modification prevention), availability (timely reliable access).

Non-repudiation requires proof of origin binding sender to message, typically via digital signature, not symmetric MAC alone.

Defense in depth layers controls so a single failed control does not yield full compromise; each layer assumes prior layers fail.

Least privilege: subject receives only the minimum rights needed for the task, for the minimum duration.

Threat vs. vulnerability vs. risk: threat exploits, vulnerability enables, risk = likelihood x impact of exploitation.

Identification asserts an identity claim; authentication verifies it; authorization grants permissions; accountability ties actions to identities via audit.

Fail-safe defaults: a control denying access on failure (fail-closed) is preferred for confidentiality; fail-open favors availability.

Separation of duties splits a critical task across two or more subjects so no single subject can complete it alone.

Classification labels (public, internal, confidential, restricted) drive handling, retention, and encryption requirements.

CIA triad: confidentiality (disclosure prevention), integrity (unauthorized modification prevention), availability (timely reliable access).

Non-repudiation requires proof of origin binding sender to message, typically via digital signature, not symmetric MAC alone.

Defense in depth layers controls so a single failed control does not yield full compromise; each layer assumes prior layers fail.

Least privilege: subject receives only the minimum rights needed for the task, for the minimum duration.

Threat vs. vulnerability vs. risk: threat exploits, vulnerability enables, risk = likelihood x impact of exploitation.

Identification asserts an identity claim; authentication verifies it; authorization grants permissions; accountability ties actions to identities via audit.

Fail-safe defaults: a control denying access on failure (fail-closed) is preferred for confidentiality; fail-open favors availability.

Separation of duties splits a critical task across two or more subjects so no single subject can complete it alone.

Classification labels (public, internal, confidential, restricted) drive handling, retention, and encryption requirements.

CIA triad: confidentiality (disclosure prevention), integrity (unauthorized modification prevention), availability (timely reliable access).

Non-repudiation requires proof of origin binding sender to message, typically via digital signature, not symmetric MAC alone.

Defense in depth layers controls so a single failed control does not yield full compromise; each layer assumes prior layers fail.

Least privilege: subject receives only the minimum rights needed for the task, for the minimum duration.

Threat vs. vulnerability vs. risk: threat exploits, vulnerability enables, risk = likelihood x impact of exploitation.

Identification asserts an identity claim; authentication verifies it; authorization grants permissions; accountability ties actions to identities via audit.

Fail-safe defaults: a control denying access on failure (fail-closed) is preferred for confidentiality; fail-open favors availability.

Separation of duties splits a critical task across two or more subjects so no single subject can complete it alone.

Classification labels (public, internal, confidential, restricted) drive handling, retention, and encryption requirements.

CIA triad: confidentiality (disclosure prevention), integrity (unauthorized modification prevention), availability (timely reliable access).

Non-repudiation requires proof of origin binding sender to message, typically via digital signature, not symmetric MAC alone.

Defense in depth layers controls so a single failed control does not yield full compromise; each layer assumes prior layers fail.

Least privilege: subject receives only the minimum rights needed for the task, for the minimum duration.

Threat vs. vulnerability vs. risk: threat exploits, vulnerability enables, risk = likelihood x impact of exploitation.

Identification asserts an identity claim; authentication verifies it; authorization grants permissions; accountability ties actions to identities via audit.

Fail-safe defaults: a control denying access on failure (fail-closed) is preferred for confidentiality; fail-open favors availability.

Separation of duties splits a critical task across two or more subjects so no single subject can complete it alone.

Classification labels (public, internal, confidential, restricted) drive handling, retention, and encryption requirements.

CIA triad: confidentiality (disclosure prevention), integrity (unauthorized modification prevention), availability (timely reliable access).

Non-repudiation requires proof of origin binding sender to message, typically via digital signature, not symmetric MAC alone.

Defense in depth layers controls so a single failed control does not yield full compromise; each layer assumes prior layers fail.

Least privilege: subject receives only the minimum rights needed for the task, for the minimum duration.

Threat vs. vulnerability vs. risk: threat exploits, vulnerability enables, risk = likelihood x impact of exploitation.

Identification asserts an identity claim; authentication verifies it; authorization grants permissions; accountability ties actions to identities via audit.

Fail-safe defaults: a control denying access on failure (fail-closed) is preferred for confidentiality; fail-open favors availability.

Separation of duties splits a critical task across two or more subjects so no single subject can complete it alone.

Classification labels (public, internal, confidential, restricted) drive handling, retention, and encryption requirements.

CIA triad: confidentiality (disclosure prevention), integrity (unauthorized modification prevention), availability (timely reliable access).

Non-repudiation requires proof of origin binding sender to message, typically via digital signature, not symmetric MAC alone.

Defense in depth layers controls so a single failed control does not yield full compromise; each layer assumes prior layers fail.

Least privilege: subject receives only the minimum rights needed for the task, for the minimum duration.

Threat vs. vulnerability vs. risk: threat exploits, vulnerability enables, risk = likelihood x impact of exploitation.

Identification asserts an identity claim; authentication verifies it; authorization grants permissions; accountability ties actions to identities via audit.

Fail-safe defaults: a control denying access on failure (fail-closed) is preferred for confidentiality; fail-open favors availability.

Separation of duties splits a critical task across two or more subjects so no single subject can complete it alone.

Classification labels (public, internal, confidential, restricted) drive handling, retention, and encryption requirements.

CIA triad: confidentiality (disclosure prevention), integrity (unauthorized modification prevention), availability (timely reliable access).

Non-repudiation requires proof of origin binding sender to message, typically via digital signature, not symmetric MAC alone.

Defense in depth layers controls so a single failed control does not yield full compromise; each layer assumes prior layers fail.

Least privilege: subject receives only the minimum rights needed for the task, for the minimum duration.

Threat vs. vulnerability vs. risk: threat exploits, vulnerability enables, risk = likelihood x impact of exploitation.

Identification asserts an identity claim; authentication verifies it; authorization grants permissions; accountability ties actions to identities via audit.

Fail-safe defaults: a control denying access on failure (fail-closed) is preferred for confidentiality; fail-open favors availability.

Separation of duties splits a critical task across two or more subjects so no single subject can complete it alone.

Classification labels (public, internal, confidential, restricted) drive handling, retention, and encryption requirements.

CIA triad: confidentiality (disclosure prevention), integrity (unauthorized modification prevention), availability (timely reliable access).

Non-repudiation requires proof of origin binding sender to message, typically via digital signature, not symmetric MAC alone.

Defense in depth layers controls so a single failed control does not yield full compromise; each layer assumes prior layers fail.

Least privilege: subject receives only the minimum rights needed for the task, for the minimum duration.

Threat vs. vulnerability vs. risk: threat exploits, vulnerability enables, risk = likelihood x impact of exploitation.

Identification asserts an identity claim; authentication verifies it; authorization grants permissions; accountability ties actions to identities via audit.

Fail-safe defaults: a control denying access on failure (fail-closed) is preferred for confidentiality; fail-open favors availability.

Separation of duties splits a critical task across two or more subjects so no single subject can complete it alone.

Classification labels (public, internal, confidential, restricted) drive handling, retention, and encryption requirements.

CIA triad: confidentiality (disclosure prevention), integrity (unauthorized modification prevention), availability (timely reliable access).

Non-repudiation requires proof of origin binding sender to message, typically via digital signature, not symmetric MAC alone.

Defense in depth layers controls so a single failed control does not yield full compromise; each layer assumes prior layers fail.

Least privilege: subject receives only the minimum rights needed for the task, for the minimum duration.

Threat vs. vulnerability vs. risk: threat exploits, vulnerability enables, risk = likelihood x impact of exploitation.

Identification asserts an identity claim; authentication verifies it; authorization grants permissions; accountability ties actions to identities via audit.

Fail-safe defaults: a control denying access on failure (fail-closed) is preferred for confidentiality; fail-open favors availability.

Separation of duties splits a critical task across two or more subjects so no single subject can complete it alone.

Classification labels (public, internal, confidential, restricted) drive handling, retention, and encryption requirements.

CIA triad: confidentiality (disclosure prevention), integrity (unauthorized modification prevention), availability (timely reliable access).

Non-repudiation requires proof of origin binding sender to message, typically via digital signature, not symmetric MAC alone.

Defense in depth layers controls so a single failed control does not yield full compromise; each layer assumes prior layers fail.

Least privilege: subject receives only the minimum rights needed for the task, for the minimum duration.

Threat vs. vulnerability vs. risk: threat exploits, vulnerability enables, risk = likelihood x impact of exploitation.

Identification asserts an identity claim; authentication verifies it; authorization grants permissions; accountability ties actions to identities via audit.

Fail-safe defaults: a control denying access on failure (fail-closed) is preferred for confidentiality; fail-open favors availability.

Separation of duties splits a critical task across two or more subjects so no single subject can complete it alone.

Classification labels (public, internal, confidential, restricted) drive handling, retention, and encryption requirements.

CIA triad: confidentiality (disclosure prevention), integrity (unauthorized modification prevention), availability (timely reliable access).

Non-repudiation requires proof of origin binding sender to message, typically via digital signature, not symmetric MAC alone.

Defense in depth layers controls so a single failed control does not yield full compromise; each layer assumes prior layers fail.

Least privilege: subject receives only the minimum rights needed for the task, for the minimum duration.

Threat vs. vulnerability vs. risk: threat exploits, vulnerability enables, risk = likelihood x impact of exploitation.

Identification asserts an identity claim; authentication verifies it; authorization grants permissions; accountability ties actions to identities via audit.

Fail-safe defaults: a control denying access on failure (fail-closed) is preferred for confidentiality; fail-open favors availability.

Separation of duties splits a critical task across two or more subjects so no single subject can complete it alone.

Classification labels (public, internal, confidential, restricted) drive handling, retention, and encryption requirements.

CIA triad: confidentiality (disclosure prevention), integrity (unauthorized modification prevention), availability (timely reliable access).

Non-repudiation requires proof of origin binding sender to message, typically via digital signature, not symmetric MAC alone.

Defense in depth layers controls so a single failed control does not yield full compromise; each layer assumes prior layers fail.

Least privilege: subject receives only the minimum rights needed for the task, for the minimum duration.

Threat vs. vulnerability vs. risk: threat exploits, vulnerability enables, risk = likelihood x impact of exploitation.

Identification asserts an identity claim; authentication verifies it; authorization grants permissions; accountability ties actions to identities via audit.

Fail-safe defaults: a control denying access on failure (fail-closed) is preferred for confidentiality; fail-open favors availability.

Separation of duties splits a critical task across two or more subjects so no single subject can complete it alone.

Classification labels (public, internal, confidential, restricted) drive handling, retention, and encryption requirements.

CIA triad: confidentiality (disclosure prevention), integrity (unauthorized modification prevention), availability (timely reliable access).

Non-repudiation requires proof of origin binding sender to message, typically via digital signature, not symmetric MAC alone.

Defense in depth layers controls so a single failed control does not yield full compromise; each layer assumes prior layers fail.

Least privilege: subject receives only the minimum rights needed for the task, for the minimum duration.

Threat vs. vulnerability vs. risk: threat exploits, vulnerability enables, risk = likelihood x impact of exploitation.

Identification asserts an identity claim; authentication verifies it; authorization grants permissions; accountability ties actions to identities via audit.

Fail-safe defaults: a control denying access on failure (fail-closed) is preferred for confidentiality; fail-open favors availability.

Separation of duties splits a critical task across two or more subjects so no single subject can complete it alone.

Classification labels (public, internal, confidential, restricted) drive handling, retention, and encryption requirements.

CIA triad: confidentiality (disclosure prevention), integrity (unauthorized modification prevention), availability (timely reliable access).

Non-repudiation requires proof of origin binding sender to message, typically via digital signature, not symmetric MAC alone.

Defense in depth layers controls so a single failed control does not yield full compromise; each layer assumes prior layers fail.

Least privilege: subject receives only the minimum rights needed for the task, for the minimum duration.

Threat vs. vulnerability vs. risk: threat exploits, vulnerability enables, risk = likelihood x impact of exploitation.

Identification asserts an identity claim; authentication verifies it; authorization grants permissions; accountability ties actions to identities via audit.

Fail-safe defaults: a control denying access on failure (fail-closed) is preferred for confidentiality; fail-open favors availability.

Separation of duties splits a critical task across two or more subjects so no single subject can complete it alone.

Classification labels (public, internal, confidential, restricted) drive handling, retention, and encryption requirements.

CIA triad: confidentiality (disclosure prevention), integrity (unauthorized modification prevention), availability (timely reliable access).

Non-repudiation requires proof of origin binding sender to message, typically via digital signature, not symmetric MAC alone.

Defense in depth layers controls so a single failed control does not yield full compromise; each layer assumes prior layers fail.

Least privilege: subject receives only the minimum rights needed for the task, for the minimum duration.

Threat vs. vulnerability vs. risk: threat exploits, vulnerability enables, risk = likelihood x impact of exploitation.

Identification asserts an identity claim; authentication verifies it; authorization grants permissions; accountability ties actions to identities via audit.

Fail-safe defaults: a control denying access on failure (fail-closed) is preferred for confidentiality; fail-open favors availability.

Separation of duties splits a critical task across two or more subjects so no single subject can complete it alone.

Classification labels (public, internal, confidential, restricted) drive handling, retention, and encryption requirements.

CIA triad: confidentiality (disclosure prevention), integrity (unauthorized modification prevention), availability (timely reliable access).

Non-repudiation requires proof of origin binding sender to message, typically via digital signature, not symmetric MAC alone.

Defense in depth layers controls so a single failed control does not yield full compromise; each layer assumes prior layers fail.

Least privilege: subject receives only the minimum rights needed for the task, for the minimum duration.

Threat vs. vulnerability vs. risk: threat exploits, vulnerability enables, risk = likelihood x impact of exploitation.

Identification asserts an identity claim; authentication verifies it; authorization grants permissions; accountability ties actions to identities via audit.

Fail-safe defaults: a control denying access on failure (fail-closed) is preferred for confidentiality; fail-open favors availability.

Separation of duties splits a critical task across two or more subjects so no single subject can complete it alone.

Classification labels (public, internal, confidential, restricted) drive handling, retention, and encryption requirements.

CIA triad: confidentiality (disclosure prevention), integrity (unauthorized modification prevention), availability (timely reliable access).

Non-repudiation requires proof of origin binding sender to message, typically via digital signature, not symmetric MAC alone.

Defense in depth layers controls so a single failed control does not yield full compromise; each layer assumes prior layers fail.

Least privilege: subject receives only the minimum rights needed for the task, for the minimum duration.

Threat vs. vulnerability vs. risk: threat exploits, vulnerability enables, risk = likelihood x impact of exploitation.

Identification asserts an identity claim; authentication verifies it; authorization grants permissions; accountability ties actions to identities via audit.

Fail-safe defaults: a control denying access on failure (fail-closed) is preferred for confidentiality; fail-open favors availability.

Separation of duties splits a critical task across two or more subjects so no single subject can complete it alone.

Classification labels (public, internal, confidential, restricted) drive handling, retention, and encryption requirements.

CIA triad: confidentiality (disclosure prevention), integrity (unauthorized modification prevention), availability (timely reliable access).

Non-repudiation requires proof of origin binding sender to message, typically via digital signature, not symmetric MAC alone.

Defense in depth layers controls so a single failed control does not yield full compromise; each layer assumes prior layers fail.

Least privilege: subject receives only the minimum rights needed for the task, for the minimum duration.

Threat vs. vulnerability vs. risk: threat exploits, vulnerability enables, risk = likelihood x impact of exploitation.

Identification asserts an identity claim; authentication verifies it; authorization grants permissions; accountability ties actions to identities via audit.

Fail-safe defaults: a control denying access on failure (fail-closed) is preferred for confidentiality; fail-open favors availability.

Separation of duties splits a critical task across two or more subjects so no single subject can complete it alone.

Classification labels (public, internal, confidential, restricted) drive handling, retention, and encryption requirements.

CIA triad: confidentiality (disclosure prevention), integrity (unauthorized modification prevention), availability (timely reliable access).

Non-repudiation requires proof of origin binding sender to message, typically via digital signature, not symmetric MAC alone.

Defense in depth layers controls so a single failed control does not yield full compromise; each layer assumes prior layers fail.

Least privilege: subject receives only the minimum rights needed for the task, for the minimum duration.

Threat vs. vulnerability vs. risk: threat exploits, vulnerability enables, risk = likelihood x impact of exploitation.

Identification asserts an identity claim; authentication verifies it; authorization grants permissions; accountability ties actions to identities via audit.

Fail-safe defaults: a control denying access on failure (fail-closed) is preferred for confidentiality; fail-open favors availability.

Separation of duties splits a critical task across two or more subjects so no single subject can complete it alone.

Classification labels (public, internal, confidential, restricted) drive handling, retention, and encryption requirements.