Domain 5 · Lesson 1 of 5

Authentication Factors & MFA

Nhân tố Xác thực & Đa yếu tố

The 4 A's — In Sequence

1. Identification

Claim identity (username, card swipe)

→

2. Authentication

Prove the claim (password, OTP, biometric)

→

3. Authorization

Grant permitted access (RBAC, ABAC)

→

4. Accountability

Audit trail (logs, session recording)

Three Authentication Factor Types

Type Tiếng Việt Examples Weakness
Type 1: Know Điều bạn biết Password, PIN, passphrase, security question Phishing, keylogging, shoulder surfing
Type 2: Have Điều bạn có SMS OTP, TOTP app, hardware token, smart card, YubiKey SIM swap (SMS), theft
Type 3: Are Điều bạn là Fingerprint, face, iris, retina, voice, vein Spoofing (need liveness detection)

MFA RULE: Must combine 2+ DIFFERENT factor types

  • • Phone number + SMS OTP = ONLY Type 2 (not true MFA — same factor type twice)
  • • Password + OTP = Type 1 + Type 2 = true MFA ✓
  • • Partner E MPIN + biometric = Type 1 + Type 3 = true MFA ✓

Biometric Accuracy Metrics

FAR — False Accept Rate

Wrong person gets in. Security risk. Also called Type II error or False Positive in security context.

FRR — False Reject Rate

Right person locked out. Usability risk. Also called Type I error.

CER / EER — Crossover Error Rate

Point where FAR = FRR. Lower CER = more accurate system.

FAR and FRR trade off against each other — tightening sensitivity reduces FAR but increases FRR. CER is the single number used to compare biometric systems: a system with CER 1% is more accurate than one with CER 5%.

NIST SP 800-63A — Identity Proofing Levels

Identity proofing = verifying that a person IS who they claim to be before issuing credentials. Three levels of assurance:

LevelNameProcessUse Case
IAL1Identity Assurance Level 1Self-assertion only — no identity verification. Applicant states their own identity.Low-risk accounts, anonymous services
IAL2Identity Assurance Level 2Remote or in-person identity proofing with identity evidence (government ID). Document validation + biometric comparison.Financial accounts, most government services
IAL3Identity Assurance Level 3In-person identity proofing by trained representative. Physical document inspection. Highest assurance.High-value government clearances, banking

Non-Person Entities (Service Accounts, API Keys, IoT Devices)

Authentication extends beyond humans: Service accounts = authenticated via client certificates or OIDC federation (not passwords); API keys = Type 2 factor (something you HAVE) — must be rotated and stored in Vault; IoT devices = device certificates, hardware TPM attestation. All non-person entities should have unique identities with least privilege.

FinTech Company X: Platform C microservices use mTLS client certificates for service-to-service auth (Type 3 equivalent for services). Partner D B2B partner uses HMAC-signed requests (API key equivalent). All secrets managed in HashiCorp Vault with auto-rotation.

Password Best Practices — NIST SP 800-63B

Recommended ✓

  • ✓ Prioritize length over complexity (passphrases)
  • ✓ Check against breached password lists (HaveIBeenPwned)
  • ✓ Allow paste in password fields
  • ✓ Store with bcrypt / Argon2 / scrypt (slow + salted)

Avoid ✗

  • ✗ Mandatory periodic changes unless compromised
  • ✗ Arbitrary complexity rules (special chars, caps)
  • ✗ MD5 or SHA-1 for password hashing (fast = bad)
  • ✗ Blocking paste (forces weaker passwords)

Key Terms

Authentication MFA Type 1/2/3 Factor FAR FRR CER Biometrics Liveness Detection OTP TOTP FIDO2/WebAuthn bcrypt Salt
Exam Tips
  1. MFA requires DIFFERENT factor types — same type twice is NOT MFA (phone number + SMS OTP = only Type 2).
  2. FAR = security risk (lets wrong person in); FRR = usability risk (locks right person out).
  3. Lower CER = more accurate biometric system.
  4. NIST 800-63B recommends length over complexity — do NOT force periodic changes unless compromised.
  5. FIDO2/WebAuthn = phishing-resistant MFA — asymmetric key pair on device, private key never leaves device.
Work Application — FinTech Company X

Partner E biometric MFA: MPIN (Type 1) + biometric face scan (Type 3) = true MFA. eKYC Vendor liveness detection prevents photo spoofing and video replay attacks — essential because face data can be photographed.

OTP rate limiting: OTP capped at 5/hr per phone number prevents Type 2 brute force. Anti-enumeration: OTP always sent regardless of account existence — prevents attackers from discovering valid accounts by observing different responses.

Recommendation: Upgrade Partner C from SMS OTP (Type 2, SIM-swap risk) to TOTP app option (Authenticator app) for users who prefer stronger Type 2 authentication without SIM-swap exposure.

Practice Quiz

Q1. A user logs in with their phone number and then receives an SMS OTP. Is this true MFA?

▼ Reveal Answer
No — this is NOT true MFA. Phone number (something you have) + SMS OTP (something you have) = only Type 2 factor used twice.
MFA requires combining two or more DIFFERENT factor types. Phone + SMS is both "something you have" (Type 2). To make it true MFA, add a password (Type 1: something you know) or biometric (Type 3: something you are).

Q2. Partner E uses MPIN plus biometric face scan. Which factor types are used, and is this true MFA?

▼ Reveal Answer
MPIN = Type 1 (something you know). Biometric face = Type 3 (something you are). Yes — this is true MFA combining two different factor types.
MPIN is a Personal Identification Number — knowledge factor (Type 1). Face biometric is an inherence factor (Type 3). Two different types = true multi-factor authentication. eKYC Vendor liveness detection further prevents spoofing attacks against the biometric factor.

Q3. Which biometric error rate represents a security risk — FAR or FRR?

▼ Reveal Answer
FAR (False Accept Rate) is the security risk — it means an unauthorized person is incorrectly granted access.
FAR = false positive from a security perspective: wrong person gets in. FRR = false negative: right person is locked out (usability problem). On CISSP, FAR is the primary security concern. CER is the crossover point where FAR equals FRR — a lower CER means the system is more accurate overall.

Q4. What does CER measure, and what does a lower CER indicate about a biometric system?

▼ Reveal Answer
CER (Crossover Error Rate) is the point where FAR equals FRR. A lower CER means the system is more accurate overall.
CER is used to compare biometric systems on a single number. A system with CER 1% makes fewer errors overall than one with CER 5%, because the crossover point (where false accepts equal false rejects) is lower. When comparing two biometric solutions, always choose the one with lower CER.

Q5. Liveness detection — what specific attack does it prevent in the context of biometric authentication?

▼ Reveal Answer
Liveness detection prevents spoofing attacks — using a photograph, printed image, or video replay to fool a facial recognition or fingerprint scanner into granting access.
Without liveness detection, an attacker could defeat face recognition by holding up a photo. Liveness detection verifies the biometric sample is from a live human (blink detection, 3D depth sensing, micro-expressions). eKYC Vendor implements this for Partner E. For CISSP: Type 3 factors need liveness detection to prevent spoofing — this is their primary weakness.