IAM Foundations: The Security Brain

Enterprise Identity & Access Management - Complete Theory Guide

Enhanced Edition v2.0
🧠 The Security Brain of Enterprise

Think of Identity and Access Management (IAM) as the "Security Brain" of a giant digital castle. This brain is responsible for knowing exactly who every person is, what they're allowed to do, and ensuring they don't have too much power. In modern enterprises, IAM is the foundational layer that enables Zero Trust architecture and protects against 80% of data breaches.

🔐
Authentication
Who are you?
Verify identity claims
🎯
Authorization
What can you do?
Enforce access policies
📊
Accounting
What did you do?
Audit trail & logging
🛡️
Administration
How do we manage it?
Lifecycle governance
🏢
Enterprise Reality Check: In organizations with 10,000+ employees, a single IAM system might handle 50,000+ identities (employees, contractors, service accounts, machine identities), 500+ applications, and process millions of authentication requests daily. Proper IAM design directly impacts user productivity, security posture, and compliance status.

📋 Fundamental Security Principles

🔒
Need-to-Know (Confidentiality)
Even if you're a high-level manager, you shouldn't see a file unless you specifically need it for your current task. This principle protects sensitive data from unnecessary exposure.
Example: The CEO doesn't get access to the HR system's salary database unless they're actively reviewing compensation.
In healthcare (HIPAA), providers can only access patient records they're actively treating. Audit logs track every access.
Least Privilege (Integrity)
You should only have the minimum "powers" that you absolutely need to do your job. No more, no less. This limits the blast radius of compromised accounts.
Example: A student can read their grades but cannot edit them. A teacher can edit grades but only for their own classes.
AWS IAM policies should use specific resource ARNs, not `Resource: "*"`. Cloud security posture management (CSPM) tools flag overly permissive policies.
🎭
Separation of Duties (SoD)
Critical tasks are split between multiple people to prevent fraud or errors. No single person should control an entire sensitive process.
Example: The person who creates a purchase order cannot be the same person who approves the payment.
ERP systems like SAP implement SoD matrices. Identity Governance tools (SailPoint, Saviynt) detect toxic access combinations automatically.
🔄
Defense in Depth
Multiple layers of security controls, so if one fails, others still protect. Never rely on a single security mechanism.
Example: Password + MFA + Device Trust + Network Location = Multiple barriers.
Modern Zero Trust architectures implement continuous verification at identity, device, network, application, and data layers.
👁️
Zero Trust
"Never trust, always verify" - assume breach and verify every request regardless of source. The network perimeter is no longer the security boundary.
Example: Even inside the corporate network, every access request is authenticated and authorized.
NIST SP 800-207 defines Zero Trust Architecture. Major implementations include Google BeyondCorp and Microsoft Zero Trust.
⏱️
Time-Based Access
Access rights that automatically expire after a set period. Standing privileges are the enemy of security.
Example: Contractor accounts automatically disable after project end date.
Azure PIM (Privileged Identity Management) enables just-in-time activation with configurable TTLs from 30 minutes to 24 hours.

🎮 Access Control Models

💡
Key Insight: Organizations typically use a combination of these models, not just one. The choice depends on organizational size, compliance requirements, and operational complexity. Most enterprises start with RBAC and evolve toward ABAC for fine-grained control.
👥
RBAC (Role-Based Access Control)
Access based on your job title or role in the organization. Instead of giving permissions to "Alex," give permissions to the "Editor" role. This simplifies management at scale.
Example: All "Developers" can access the code repository. All "Accountants" can access the financial system.
Implemented in Active Directory groups, AWS IAM roles, Kubernetes RBAC. NIST defines RBAC0-RBAC3 maturity levels.
🏷️
ABAC (Attribute-Based Access Control)
The smartest model - looks at the context: WHO you are, WHAT you're accessing, WHEN, WHERE, and HOW. Policies are written using attributes, not static roles.
Example: You can access the server (Role), but ONLY between 9 AM-5 PM (Time) from a company laptop (Device) in the office (Location).
AWS IAM Conditions, Azure Conditional Access, and XACML (eXtensible Access Control Markup Language) enable ABAC at scale.
🏛️
MAC (Mandatory Access Control)
Government/military style - strict levels like "Top Secret," "Secret," "Confidential," "Unclassified." Classification is mandatory, not discretionary.
Example: Someone with "Secret" clearance cannot access "Top Secret" documents, period.
Implemented via SELinux, AppArmor, Windows MIC (Mandatory Integrity Control). Required for government contracts (FedRAMP, IL4-IL6).
📝
DAC (Discretionary Access Control)
The owner of a resource decides who can access it. Simple but can lead to inconsistent security when scaled.
Example: You create a document and choose to share it with specific teammates.
File system permissions (chmod), SharePoint sharing, Google Drive. Risk: users may over-share sensitive data.
📊
ReBAC (Relationship-Based Access Control)
Access is determined by the relationship between subjects and objects in a graph. Modern approach used by Google Zanzibar and similar systems.
Example: You can edit a document if you're a member of a team that owns the folder containing it.
Google Zanzibar, Ory Keto, SpiceDB, Authzed. Enables complex permission hierarchies in SaaS applications.
⚖️
Policy-Based Access Control (PBAC)
Centralized policy engine evaluates access requests against defined rules. Combines elements of RBAC and ABAC with externalized authorization.
Example: A central policy states "Only verified humans from approved countries can access PII during business hours."
Open Policy Agent (OPA), AWS Cedar, Styra DAS. Enables "policy as code" with version control and testing.

Access Control Model Comparison

Model Best For Complexity Flexibility Scalability
RBAC Medium to large organizations with defined roles Medium Moderate High
ABAC Complex environments needing fine-grained control High Very High Medium
MAC Government, military, high-security environments High Very Low Medium
DAC Small teams, collaborative environments Low High Low
ReBAC Complex SaaS apps with hierarchical resources High Very High Very High
PBAC Cloud-native, microservices architectures Medium-High High Very High

👑 Privileged Access Management (PAM)

The Crown Jewels of Security

Some accounts are more dangerous than others. Privileged accounts (like "Admins") have the power to delete the whole system, exfiltrate all data, or cover their tracks. PAM keeps them on a tight leash with vaulting, session control, and just-in-time elevation.

⚠️
The Privileged Account Problem: 80% of security breaches involve privileged credentials. A compromised admin account gives attackers "god mode" access to your entire infrastructure. Standing privileges (always-on admin access) are the #1 target for attackers.
Just-In-Time (JIT) Access
Admins don't walk around with "God Mode" on all day. They "check out" privileges only when needed, for a specific duration, with full justification and audit trail.
Example: Need to fix a server? Request admin access for 30 minutes. After that, it automatically expires.
CyberArk PAS, HashiCorp Vault, Azure PIM, AWS IAM Identity Center. Typical JIT windows: 1-4 hours max.
📹
Session Recording & Monitoring
Every action taken with privileged access is recorded like a security camera. Video-like playback enables forensic investigation and compliance evidence.
Example: Admin sessions are recorded so you can replay exactly what commands were run if something goes wrong.
CyberArk PSM, BeyondTrust, Teleport. Sessions include keystroke logging, command capture, and screen recording.
🔑
Credential Vaulting
Passwords for privileged accounts are stored in a secure vault and rotated automatically. Humans never see or know the actual passwords.
Example: Database admin password changes every hour, and only the vault knows the current one.
Rotation schedules: service accounts (30-90 days), shared accounts (daily), emergency accounts (after each use).
🚨
Privilege Escalation Detection
Detect when a hacker tries to "level up" from a regular user to admin. Monitor for unusual elevation patterns and lateral movement.
Example: Alert when a normal user account suddenly tries to run admin commands or access privileged systems.
SIEM correlation rules, UEBA (User Entity Behavior Analytics), EDR privilege escalation detection.
Approval Workflows
High-risk actions require approval from multiple people. Automated workflows ensure proper authorization without slowing operations.
Example: Deleting a production database requires approval from both IT Manager and Security team.
ServiceNow integration, Slack/Teams approval bots, time-based auto-approval for on-call scenarios.
🔄
Break-Glass Procedures
Emergency access process for critical situations, with heavy auditing. Used when normal approval workflows would cause unacceptable delay.
Example: System is down at 3 AM? Use break-glass account, but expect a security review tomorrow.
Break-glass accounts must be: monitored 24/7, auto-rotated after use, trigger immediate SOC alerts, require post-incident review.

🔄 The IAM Lifecycle

👶
Provisioning
🎯
Authentication
🔐
Authorization
📊
Monitoring
🔄
Re-certification
💀
Deprovisioning
👶
Provisioning (Joiner)
Creating a new digital identity when someone joins. Includes account creation, initial access assignment, and credential issuance.
HR-driven provisioning via SCIM, automated birthright access based on role, Day 1 productivity SLAs (accounts ready before start date).
🔄
Movers (Role Changes)
When someone changes roles, their access must be adjusted. Old access removed, new access granted - preventing access accumulation.
Most dangerous phase - users often keep old access. IGA tools automate access recertification on role change events.
💀
Deprovisioning (Leaver)
Removing all access when someone leaves. The most critical and often neglected phase of the identity lifecycle.
Target SLA: high-risk accounts disabled within 1 hour of termination. HR integration enables automated triggering.
⚠️
Critical: Deprovisioning is the most important step! When someone is fired or leaves, their digital identity must be "killed" instantly to prevent revenge attacks or data theft. Studies show 89% of ex-employees retain access to at least one corporate application after leaving.

🔐 Authentication Protocols Deep Dive

💡
Why This Matters: Understanding authentication protocols is essential for designing SSO solutions, integrating applications, and troubleshooting authentication failures. These protocols are the language IAM systems speak.
Modern Standard
OpenID Connect (OIDC)

OIDC is an identity layer built on top of OAuth 2.0. It provides authentication (who you are) while OAuth 2.0 provides authorization (what you can access). OIDC is the de facto standard for modern web and mobile applications.

1
User clicks "Login"
2
Redirect to IdP
3
User authenticates
4
Authorization Code
5
Exchange for Tokens
6
ID Token + Access Token
OIDC ID Token (JWT) Structure
{
  "iss": "https://idp.example.com",     // Issuer - who created this token
  "sub": "user-12345",                   // Subject - unique user identifier
  "aud": "my-app-client-id",             // Audience - intended recipient
  "exp": 1735689600,                     // Expiration time (Unix timestamp)
  "iat": 1735686000,                     // Issued at time
  "auth_time": 1735685900,               // When user actually authenticated
  "nonce": "abc123",                     // Replay attack prevention
  "email": "user@example.com",           // Standard claim
  "groups": ["admins", "developers"]     // Custom claim for RBAC
}
Enterprise Standard
SAML 2.0 (Security Assertion Markup Language)

SAML is the legacy enterprise standard for SSO, using XML-based assertions. Still widely used for enterprise applications that predate OIDC adoption. Most enterprise IdPs support both SAML and OIDC.

1
Access SP
2
SAML Request
3
IdP Authentication
4
SAML Response
5
Assertion Validated

OIDC vs SAML Comparison

Aspect OIDC SAML 2.0
Format JSON (JWT) XML
Best For Modern web, mobile, APIs Enterprise legacy apps
Token Size Compact (~1KB) Large (~10KB+)
Complexity Lower Higher
Mobile Support Native Challenging
On-Premises
Kerberos

Kerberos is the authentication protocol used by Active Directory. It uses tickets instead of passwords, providing SSO for Windows domain environments. Understanding Kerberos is essential for hybrid identity scenarios.

1
AS-REQ to KDC
2
TGT Issued
3
TGS-REQ
4
Service Ticket
5
Access Service
⚠️
Security Considerations: Kerberos attacks include Golden Ticket (forged TGT), Silver Ticket (forged service ticket), Kerberoasting (cracking service account passwords), and Pass-the-Ticket. Modern AD security requires monitoring for these attack patterns.

🛡️ Zero Trust Architecture

Never Trust, Always Verify

Zero Trust is a security model that eliminates implicit trust. Every access request is fully authenticated, authorized, and encrypted before granting access—regardless of where the request originates or what resource it accesses.

📋
NIST SP 800-207: The authoritative Zero Trust Architecture standard. Defines ZTA tenets, deployment models, and migration strategies. Required reading for enterprise architects.

The Five Pillars of Zero Trust

🆔 Identity

Verify and secure every identity with strong authentication before granting access.

🔐
Strong MFA
👤
Identity Proofing
📊
Risk-Based Auth
🤖
Machine Identity

💻 Devices

Ensure devices meet security requirements before granting access to resources.

Device Compliance
🔒
Endpoint Security
📱
MDM/MAM
🏥
Health Attestation

🌐 Network

Segment networks and enforce micro-perimeters around sensitive resources.

🧱
Micro-Segmentation
🔐
Encrypted Traffic
🚫
No Implicit Trust
🔍
Traffic Inspection

📱 Applications

Secure and monitor all applications, including shadow IT discovery.

🔐
App-Level Auth
👻
Shadow IT Discovery
🛡️
CASB Integration
📊
API Security

💾 Data

Classify, label, and encrypt data. Apply data-centric security controls.

🏷️
Data Classification
🔐
Encryption at Rest
🚫
DLP Policies
🔍
Data Discovery

🔐 Multi-Factor Authentication (MFA)

Impact: MFA blocks 99.9% of automated attacks and 96% of phishing attacks. It's the single most effective security control you can implement.

Authentication Factors

🧠
Something You Know
Knowledge factors: passwords, PINs, security questions.
Weakest factor - vulnerable to phishing, credential stuffing, and social engineering.
📱
Something You Have
Possession factors: hardware tokens, mobile phones, smart cards.
Stronger but can be lost, stolen, or SIM-swapped. Hardware tokens most secure.
👆
Something You Are
Biometric factors: fingerprint, face recognition, iris scan.
Convenient but cannot be changed if compromised. Privacy considerations apply.

MFA Methods Comparison

A+
🔑 FIDO2/WebAuthn (Passkeys)

Phishing-resistant, passwordless authentication using public key cryptography.

  • Phishing-resistant by design
  • No shared secrets
  • Works cross-platform
  • Requires modern browsers
  • Account recovery complexity
A
🔐 Hardware Security Keys

Physical USB/NFC devices like YubiKey that generate cryptographic proofs.

  • Highly phishing-resistant
  • No battery required
  • Tamper-resistant
  • Can be lost/forgotten
  • Per-device cost ($25-50)
B
📱 Authenticator Apps (TOTP)

Time-based one-time passwords from apps like Google Authenticator, Authy.

  • Free and widely supported
  • Works offline
  • No hardware cost
  • Phishable (real-time relay)
  • Device loss = lockout
B-
📲 Push Notifications

Approve/deny prompts sent to mobile device (Microsoft Authenticator, Duo).

  • User-friendly UX
  • Shows context (location, app)
  • Fast approval
  • MFA fatigue attacks
  • Requires internet
C
📧 Email OTP

One-time codes sent via email.

  • No app installation
  • Universal access
  • Email account compromise = MFA bypass
  • Delayed delivery
  • Easily phished
D
📱 SMS OTP

One-time codes sent via text message. Deprecated by NIST for high-value accounts.

  • Universal phone support
  • No app needed
  • SIM swapping attacks
  • SS7 vulnerabilities
  • NIST deprecated

🚨 Real-World Breach Case Studies

📚
Learning from Failures: Every major breach teaches us something about IAM weaknesses. These case studies illustrate why proper identity controls are non-negotiable in enterprise environments.
Critical
SolarWinds Supply Chain Attack (2020)
18,000+
Organizations Affected
9 months
Undetected Duration
$100M+
Remediation Cost
⚡ Attack Vector

Attackers compromised SolarWinds' build system, inserting malicious code into Orion software updates. Once deployed, the SUNBURST backdoor used stolen SAML signing certificates to forge authentication tokens, gaining persistent access to victim networks without triggering alerts.

🔓 IAM Failures
  • ❌ Compromised SAML signing keys enabled Golden SAML attacks
  • ❌ Service accounts had excessive privileges
  • ❌ Lack of monitoring for SAML token anomalies
  • ❌ Federated identity trust too broad
✅ Mitigation Controls
  • Rotate SAML/OIDC signing certificates regularly and monitor for unauthorized use
  • Implement certificate pinning for critical federation trusts
  • Monitor for impossible travel and token replay attacks
  • Apply Zero Trust principles to service accounts
  • Enable Azure AD Conditional Access with continuous access evaluation
Critical
Uber Data Breach (2022)
57M
Records Exposed
MFA Fatigue
Attack Method
$148M
Settlement Cost
⚡ Attack Vector

A teenage hacker purchased stolen credentials from the dark web, then bombarded an Uber contractor with MFA push notifications until they approved one (MFA fatigue). The attacker then used social engineering via Slack to obtain elevated privileges, eventually accessing AWS, GCP, and internal admin tools.

🔓 IAM Failures
  • ❌ Push notification MFA susceptible to fatigue attacks
  • ❌ Hardcoded credentials in PowerShell scripts on network shares
  • ❌ PAM credentials stored in plaintext
  • ❌ Excessive trust in Slack for privilege escalation
✅ Mitigation Controls
  • Implement phishing-resistant MFA (FIDO2/WebAuthn) for privileged users
  • Use number matching or additional context in push notifications
  • Never store credentials in code or scripts - use secrets management
  • Implement PAM solutions with credential vaulting and rotation
  • Require out-of-band verification for privilege escalation requests
High
Okta Support System Breach (2023)
134
Customers Impacted
HAR Files
Data Exposed
14 days
Detection Time
⚡ Attack Vector

Attackers gained access to Okta's customer support system through a compromised service account. They exfiltrated HAR files (HTTP Archive) containing session tokens, then used these tokens to impersonate legitimate users at downstream customers like BeyondTrust, Cloudflare, and 1Password.

🔓 IAM Failures
  • ❌ Service account credentials exposed on personal device
  • ❌ Session tokens in support files not sanitized
  • ❌ Insufficient monitoring of support system access patterns
  • ❌ Token binding not enforced
✅ Mitigation Controls
  • Implement session token binding to device/IP
  • Sanitize sensitive data from support artifacts before storage
  • Enforce strict device compliance for service accounts
  • Monitor for session token usage from unexpected sources
  • Implement continuous access evaluation (CAE)

📋 Compliance Framework Mapping

🏛️
Why This Matters: IAM controls are foundational to virtually every compliance framework. Understanding how IAM maps to specific controls helps justify investments and ensures audit readiness.
SOC 2 Type II

Trust Services Criteria for service organizations. Focus on Security, Availability, Processing Integrity, Confidentiality, and Privacy.

CC6.1 Logical access security - user authentication
CC6.2 Access provisioning and deprovisioning
CC6.3 Role-based access and least privilege
CC6.6 Privileged access controls
ISO 27001:2022

International standard for Information Security Management Systems (ISMS). Annex A contains 93 controls.

A.5.15 Access control policy
A.5.16 Identity management
A.5.17 Authentication information
A.5.18 Access rights provisioning
A.8.2 Privileged access rights
NIST CSF 2.0

Cybersecurity Framework with six functions: Govern, Identify, Protect, Detect, Respond, Recover.

PR.AA Identity Management & Access Control
PR.AA-01 Identities and credentials managed
PR.AA-02 Identities proofed and bound
PR.AA-05 Access permissions managed
PCI DSS 4.0

Payment Card Industry Data Security Standard. Required for any organization handling cardholder data.

Req 7 Restrict access by business need-to-know
Req 8 Identify users and authenticate access
8.3.1 MFA for all CDE access
8.6 Manage service/application accounts
HIPAA Security Rule

Health Insurance Portability and Accountability Act. Protects electronic Protected Health Information (ePHI).

§164.312(a) Access control - unique user identification
§164.312(d) Person or entity authentication
§164.308(a)(4) Access authorization and modification
CIS Controls v8

Center for Internet Security prioritized security actions. 18 control families with implementation groups.

CIS 5 Account Management
CIS 6 Access Control Management
5.3 Disable dormant accounts
6.5 Require MFA for admin access

🎮 Test Your Security Brain!

💡
Challenge: You're the Lead Security Architect for a major video game company. Apply the right security concept to each scenario!
1
The "God Mode" Problem

The Lead Developer needs to update the game's main server. This requires "Superuser" powers that can delete the entire player database. You don't want them to have this power active while they're just checking email or chatting on Discord.

❓ Which security concept should you apply?
💡 Hint: Think about temporary access that expires automatically...

✅ Answer: Just-In-Time (JIT) Access / PAM

The developer should request elevated privileges only when needed, receive them for a limited time (like 30 minutes), and have them automatically revoked when the time expires. All actions during this period should be logged and monitored.

Implementation: Use a PAM solution where the developer "checks out" admin credentials from a vault, which automatically expire and rotate. Require approval for high-risk operations.

2
The "Remote Sneak" Attack

A hacker stole the password of a Customer Support rep. The hacker is trying to log in at 3:00 AM from a country where the company doesn't even have offices. Usually, this rep only logs in from Los Angeles during the day.

❓ Which Access Model would automatically block this?
💡 Hint: It's not just about the role; it's about the context...

✅ Answer: ABAC (Attribute-Based Access Control)

ABAC evaluates multiple attributes: WHO (support rep), WHEN (3 AM vs business hours), WHERE (foreign country vs Los Angeles), and WHAT (trying to access customer data). The unusual time and location would trigger automatic denial.

Additional Defense: This should also trigger alerts for potential account compromise and require step-up MFA re-authentication. Implement impossible travel detection in your SIEM.

3
The "Fake Refund" Fraud

You want to make sure an employee can't secretly send $1,000 refunds to their own bank account. You decide that the person who requests a refund cannot be the same person who approves and sends the payment.

❓ What security principle is this?
💡 Hint: It's about splitting power between people...

✅ Answer: Segregation/Separation of Duties (SoD)

By requiring different people to initiate and approve refunds, you prevent a single person from committing fraud. This is a classic implementation of SoD.

Best Practice: Implement a workflow where: Person A requests refund → Person B reviews and approves → System automatically processes (or Person C executes). Use IGA tools to detect toxic access combinations.

4
The "MFA Fatigue" Attack

An attacker has obtained valid credentials for one of your developers. They start sending dozens of MFA push notifications to the developer's phone at 2 AM, hoping they'll approve one just to make it stop.

❓ What controls would prevent this attack?
💡 Hint: Think about what makes push notifications vulnerable...

✅ Answer: Number Matching + Phishing-Resistant MFA

Several controls mitigate MFA fatigue:

  • Number Matching: User must enter a number displayed on screen, not just tap approve
  • FIDO2/WebAuthn: Phishing-resistant, no push notifications to spam
  • Rate Limiting: Block after X failed attempts
  • Context Display: Show location/device so user recognizes suspicious requests
  • Alerting: Notify security team of multiple MFA denials

🎯 Technical Interview Preparation

💼
Resume-Ready Skills: After mastering this guide, you can confidently discuss IAM fundamentals, access control models, PAM concepts, Zero Trust architecture, and compliance frameworks in technical interviews.
Q1: How would you design JIT access for a multi-account AWS environment? Architect

Key Points to Cover:

  • Use AWS IAM Identity Center (formerly SSO) with permission sets
  • Implement time-bound role assumption with session duration limits
  • Integrate with PAM solution (CyberArk, HashiCorp Vault) for approval workflows
  • Use AWS Organizations SCPs to enforce maximum session duration
  • Implement CloudTrail logging for all AssumeRole events
  • Consider AWS Secrets Manager for credential rotation
Q2: What are the trade-offs between RBAC and ABAC in a large enterprise? Senior

RBAC Advantages:

  • Easier to understand and audit
  • Maps well to organizational structures
  • Lower implementation complexity
  • Better tooling support

ABAC Advantages:

  • Fine-grained, context-aware decisions
  • Scales without role explosion
  • Supports dynamic policies
  • Better for complex environments

Recommendation: Start with RBAC for baseline access, layer ABAC for fine-grained controls. Most enterprises use a hybrid approach.

Q3: Explain how you would implement Zero Trust for a remote workforce. Architect

Zero Trust Implementation Pillars:

  • Identity: Strong MFA (FIDO2), risk-based authentication, identity proofing
  • Device: MDM/MAM enrollment, device health attestation, compliance checks
  • Network: ZTNA instead of VPN, micro-segmentation, encrypted traffic
  • Application: Per-app access policies, CASB integration, shadow IT discovery
  • Data: Classification, DLP, encryption at rest and in transit

Implementation Approach:

  • Phase 1: Strong identity foundation (MFA, SSO consolidation)
  • Phase 2: Device trust and compliance
  • Phase 3: Application access modernization
  • Phase 4: Network segmentation
  • Phase 5: Data-centric security
Q4: What's the difference between authentication and authorization? Junior

Authentication (AuthN): Verifies WHO you are

  • Proves identity through credentials (passwords, MFA, biometrics)
  • Answers "Are you who you claim to be?"
  • Examples: Login page, SSO, OIDC ID tokens

Authorization (AuthZ): Determines WHAT you can do

  • Grants or denies access to resources based on policies
  • Answers "Are you allowed to do this?"
  • Examples: RBAC roles, IAM policies, OAuth scopes

Key Point: Authentication happens first, then authorization. You must know who someone is before you can determine what they're allowed to do.

Q5: How would you detect and respond to a compromised privileged account? Senior

Detection Methods:

  • UEBA (User Entity Behavior Analytics) for anomaly detection
  • Impossible travel alerts (logins from geographically impossible locations)
  • Off-hours privileged activity monitoring
  • Failed MFA attempt spikes
  • Unusual command patterns in PAM session recordings

Response Playbook:

  • Immediate: Disable account, revoke all active sessions
  • Contain: Rotate all credentials the account had access to
  • Investigate: Review PAM logs, SIEM alerts, authentication logs
  • Remediate: Identify root cause, close attack vector
  • Recover: Re-provision account with fresh credentials after verification
✅ Knowledge Check - Are You Ready for Labs?
  • I understand the difference between authentication and authorization
  • I can explain Need-to-Know and Least Privilege principles
  • I know when to use RBAC vs ABAC vs ReBAC
  • I understand what PAM is and why it's critical
  • I can explain Just-In-Time access and credential vaulting
  • I understand Separation of Duties and why it matters
  • I know the IAM lifecycle stages (Joiner/Mover/Leaver)
  • I understand the importance of deprovisioning
  • I can explain OIDC vs SAML differences
  • I understand Zero Trust principles and the five pillars
  • I know why SMS MFA is deprecated and what alternatives exist
  • I can identify security hazards like Shadow IT and MFA fatigue
  • I understand Fail-Safe vs Fail-Secure concepts
  • I can map IAM controls to compliance frameworks
📋 Quick Reference Cheat Sheet
RBAC
Access based on your Job Title/Role
ABAC
Access based on Time, Location, Device, and more
PAM
Special locks for "Superuser" accounts
SoD
Splitting one critical task between multiple people
Least Privilege
Minimum power needed to do the job
JIT Access
Temporary privileges that auto-expire
Zero Trust
Never trust, always verify
MFA
Multiple factors to prove identity
Deprovisioning
Removing all access when someone leaves
OIDC
Modern JSON-based SSO protocol (built on OAuth2)
SAML
Legacy XML-based enterprise SSO protocol
FIDO2
Phishing-resistant passwordless authentication
🚀 Ready for Hands-On Labs!

Now that your Security Brain understands the concepts, you're ready to implement them in real environments!

Your Learning Path:

  1. IAM Foundations (You are here!)
  2. Zero Trust SSO Gateway - Build an OIDC authentication system
  3. HashiCorp Vault JIT Access - Implement privileged access management
  4. Teleport ZTNA - Deploy Zero Trust Network Access
  5. Identity Governance - Access reviews and certification campaigns
  6. Cloud IAM Deep Dive - AWS/Azure/GCP identity patterns