1 Multi-Factor Authentication
Multi-factor authentication requires evidence from more than one factor category (Module 9 hub, Section 3):
Something you know Something you have
Password + Smartphone
Something you have Something you know
Smart card + PIN
The smart card represents possession; the PIN represents knowledge — two independent categories, which is what makes it genuine MFA rather than "two things you know."
Common MFA technologies
Your verification code is 847291. Easy to deploy and familiar to users, but weak to SIM swapping, telecom account compromise, SMS interception, and social engineering — stronger than password-only, weaker than phishing-resistant methods.Approve / Deny). Much better UX than typing OTP codes — but simple push approval is vulnerable to MFA fatigue, covered next.2 MFA Fatigue, Number Matching, Step-Up & Risk-Based Auth
MFA fatigue attacks
An attacker who already knows a user's password repeatedly triggers authentication. The victim's phone shows Approve login? again and again until they approve one just to stop the notifications — also called push bombing or MFA bombing. Controls include number matching, additional context, risk-based authentication, rate limiting, device binding, and phishing-resistant authentication.
Number matching
Instead of a blind Approve button, the login screen shows a number (Enter 42 in your authenticator app) that the user must type into their device. Blind approval becomes significantly harder because the attacker's spam alone can't trick the user into matching a number they never saw.
Step-up authentication
Not every action deserves the same assurance level. A user might log into banking with password + biometric, then hit step-up authentication specifically when attempting a large transfer:
Common in banking, privileged administration, healthcare, government, payments and other high-value transactions.
Risk-based authentication
Modern identity platforms may evaluate context before deciding whether extra authentication is needed. A user who normally logs in from London on a corporate laptop at 09:00 suddenly logging in from an unknown device, on an anonymous network, at 03:00, from a different country, raises the calculated risk score — which can trigger a block or force stronger MFA. This is central to Zero Trust architectures (Module 8, Section 13).
3 Biometrics
Biometric authentication uses measurable physical or behavioural characteristics — fingerprint, facial recognition, iris scanning, voice recognition, palm recognition.
How biometrics work
Fingerprint
↓
Sensor
↓
Feature extraction
↓
Biometric template
During authentication, a new fingerprint is captured, converted into biometric data, and compared against the registered template, producing a match score:
New fingerprint
↓
Generate biometric data
↓
Compare with registered template
↓
Match score
The system decides whether the match score clears an acceptable threshold.
False acceptance and false rejection
Biometric systems are probabilistic, not a simple exact match:
Organisations tune this threshold to balance security against usability — too strict and legitimate users get locked out constantly; too loose and impostors get through.
Biometrics should usually remain local
Modern devices typically keep biometric data on the device rather than transmitting fingerprints or facial images to a remote server:
Application
↓
Request biometric verification
↓
Device secure hardware
↓
Fingerprint/Face verified locally
↓
Cryptographic key unlocked
The server receives cryptographic proof, not the user's actual fingerprint. This design is fundamental to modern passwordless authentication, covered next.
4 FIDO2 & Passkeys
FIDO2 (Fast IDentity Online) is one of the most important developments in modern authentication, combining WebAuthn (lets browsers/web apps talk to authenticators) and CTAP (lets devices talk to external authenticators like security keys) to enable strong public-key-based authentication.
Public-key authentication in FIDO2
Traditional passwords use a shared secret the server must store. FIDO2 uses asymmetric cryptography instead — the private key never leaves the authenticator.
The moving dot traces one full round trip: public key registered → server challenges → authenticator signs → server verifies. The private key never travels.
Why FIDO2 is phishing-resistant
A fake website can ask Username: / Password: and a victim can accidentally hand both over. FIDO credentials are cryptographically bound to the legitimate service — a credential registered for login.company.com cannot simply be replayed against login-company-security.example. This is what "phishing-resistant" actually means in practice, not just "harder to phish."
Hardware security keys
A physical security key (USB-A, USB-C, NFC, or Bluetooth) can act as a FIDO authenticator:
Administrator
↓
Opens cloud admin portal
↓
Enters username
↓
Security key requested
↓
Touches security key
↓
Cryptographic authentication
↓
Access granted
High-risk organisations commonly issue these to domain admins, cloud admins, security teams, developers, executives, and privileged access workstations.
User presence vs user verification
A FIDO device can therefore require possession of the authenticator plus a PIN or biometric — presence and verification combined.
Passkeys
Passkeys are a modern implementation of FIDO-based passwordless authentication, using public-key cryptography instead of a server-stored password:
Website
↓
Passkey request
↓
Phone/Laptop
↓
Face/Fingerprint/PIN
↓
Private key signs challenge
↓
Server verifies signature
The user never needs to remember or type a password.
Device-bound vs synced passkeys
Device-bound credentials stay tied to one authenticator or device — strong control, common in enterprise deployments. Synced passkeys are synchronised across a user's trusted ecosystem, improving usability since users don't lose access when replacing a device — the more common consumer approach.
Enterprise challenges with passkeys
Passkeys dramatically improve security, but organisations still have to solve: employee onboarding, lost devices, account recovery, shared workstations, privileged administrator accounts, contractors, cross-platform compatibility, device management, regulatory requirements, and help-desk recovery procedures. Passwordless technology doesn't eliminate identity management — it changes how identity is proved.
Lab Lab 1 – Walk Through an MFA Login
Create a lab user and configure MFA using any available identity platform (a free tier of Microsoft Entra ID, Google Workspace, or even a personal account with an authenticator app all work fine). Walk through and record the complete flow:
🔮 Do this, then note what you observed
1. Username entered · 2. Password validated · 3. MFA challenge initiated · 4. MFA challenge approved · 5. Authentication completed · 6. Session created.
Then repeat with: an incorrect password; a correct password + rejected MFA; a correct password + expired OTP; a correct password + successful MFA.
Reveal what to look for
Compare the resulting sign-in logs for each case — most identity platforms log the password result and the MFA result as separate events. Notice that a rejected or expired MFA challenge produces a distinctly different log entry from a wrong password, which is exactly the kind of detail a SOC analyst relies on when investigating a suspicious login (Module 8, Section 16).
Lab Lab 2 – Register and Use a Passkey
Register a passkey with any test service that supports WebAuthn (many major platforms now offer this in account security settings). Observe the registration step (key pair generated, public key registered), then authenticate again and observe the login step (challenge → biometric/PIN verification → cryptographic signature → server verification).
🔮 Predict first
Compare this to password authentication. What reusable secret did you send to the server during login?
Reveal the answer
With a properly implemented FIDO/passkey flow, the answer is essentially none — you sent a signature proving you hold the private key, not the key itself, and not a password. That's the fundamental difference from every authentication method covered so far in this module.
Lesson Outcome
You should now be able to compare MFA technologies by strength (SMS/email OTP < TOTP/push < FIDO2/passkeys), explain why MFA fatigue defeats simple push approval and how number matching helps, and explain why FIDO2/passkeys are considered phishing-resistant in a way password + OTP never fully is. Lesson 3 moves from "how a human proves who they are" to "how one system lets an application access another system on a user's behalf" — OAuth and OpenID Connect.