06 - Defender for Cloud and Endpoint¶
Microsoft Defender for Cloud (MDfC)¶
Microsoft Defender for Cloud is Microsoft's CNAPP (Cloud-Native Application Protection Platform) covering CSPM (Cloud Security Posture Management) and CWPP (Cloud Workload Protection Platform). It protects multi-cloud (Azure, AWS, GCP) and hybrid environments.
Two pillars¶
CSPM answers: What am I deploying that is misconfigured? - Continuous assessment against security baselines - Secure Score - Regulatory compliance dashboards - Recommendations with remediation guidance
CWPP answers: What threats are happening to my running workloads? - Per-workload threat detection plans (Servers, Storage, SQL, etc.) - Runtime alerts forwarded to Sentinel/XDR
CSPM Tiers¶
| Tier | Cost | Capabilities |
|---|---|---|
| Foundational CSPM | Free | Asset inventory, secure score, regulatory compliance, recommendations, attack paths (limited) |
| Defender CSPM | Per billable resource (~$5/asset/month) | Cloud Security Graph, attack path analysis, agentless secret scanning, agentless vuln scanning, sensitive data discovery, IaC scanning, container posture, EASM (preview), permissions management |
Defender Plans (CWPP)¶
| Plan | Coverage |
|---|---|
| Servers Plan 1 | MDE integration, vulnerability assessment, file integrity monitoring (preview at P1), security baselines |
| Servers Plan 2 | Adds Just-in-Time VM access, adaptive application controls, network hardening, regulatory compliance, 500MB free Sentinel ingestion for selected security data |
| Storage | Malware scanning of uploaded blobs, sensitive data threat detection, activity monitoring |
| SQL on Azure (PaaS and VMs) | Anomaly detection, SQL injection, vulnerability assessment |
| Containers | Vuln scanning of registry images and running images, runtime threat detection, K8s admission control, K8s data plane protection |
| App Service | Runtime threat detection for App Service environments |
| Key Vault | Anomalous access detection |
| Resource Manager | Suspicious ARM operations |
| DNS | Malicious DNS resolution detection |
| Open-source Relational DBs | PostgreSQL, MySQL, MariaDB threat detection |
| Cosmos DB | NoSQL anomaly detection |
| APIs (Azure API Management) | Runtime API protection |
| AI Workloads | LLM/genAI prompt injection, jailbreak, sensitive data |
Multi-Cloud Onboarding¶
AWS connector: - Created via Defender for Cloud > Environment Settings > Add environment > AWS - Deploys CloudFormation stack to AWS account creating IAM roles - Optional auto-provisioning of Arc agents for VM coverage - Discovers EC2, EKS, RDS, S3, Lambda, etc.
GCP connector: - Workload identity federation - Service account in GCP project - Cloud Build trigger for deploying agents
On-prem and other clouds: - Azure Arc enables management of non-Azure servers and Kubernetes - Once Arc-onboarded, Defender plans apply
Secure Score¶
Numeric score (0-100%) representing security posture: - Each recommendation worth a point value - Implementing recommendations or marking exemptions improves score - Group score = aggregated across subscriptions or management groups - Used as KPI for security teams; not a compliance proxy
Recommendations¶
Each recommendation has: - Severity (High/Medium/Low) - Affected resources - Remediation steps - Quick-fix automation (where supported) - Mapping to compliance frameworks
Status options: Healthy, Unhealthy, Not Applicable, Exempt.
Governance rules assign owners and due dates to recommendations to drive remediation accountability.
Regulatory Compliance¶
Built-in initiatives for: - Microsoft Cloud Security Benchmark (MCSB) - default - PCI-DSS v4 - HIPAA HITRUST - ISO 27001:2013/2022 - SOC 2 Type 2 - NIST SP 800-53, SP 800-171, CSF - CIS Microsoft Azure Foundations Benchmark - FedRAMP H/M - CMMC Level 3 - UK NHS, Australian PSPF, regional standards
Custom initiatives via Azure Policy. Compliance dashboard exports findings to PDF, supports continuous export to Sentinel and Event Hub.
Cloud Security Graph (Defender CSPM)¶
Graph database of all discovered cloud resources, identities, workloads, exposures, and relationships. Enables:
- Attack path analysis - Visualizes chains of misconfigurations that attackers could exploit (e.g., public VM with high-sev vuln + identity with admin role + access to KV with secrets)
- Cloud security explorer - Custom queries against the graph (e.g., "Internet-exposed VMs with critical CVEs and managed identities with subscription-owner role")
- Risk-prioritized recommendations based on actual exposure
Microsoft Defender for Endpoint (MDE)¶
Endpoint detection and response (EDR), antivirus (Microsoft Defender Antivirus), attack surface reduction, vulnerability management, and threat & vulnerability management.
Plans¶
| Plan | Coverage |
|---|---|
| MDE P1 | Next-gen AV, attack surface reduction, device control, web content filtering, endpoint firewall, application control |
| MDE P2 | All P1 + EDR, automated investigation and response, threat & vulnerability management, threat experts (paid), advanced hunting |
| MDE Server | MDE P2 features for server SKUs (covered by Defender for Servers Plan ½ in Defender for Cloud) |
| MDE for Business | SMB SKU bundled with Microsoft 365 Business Premium |
Onboarding methods¶
- Group Policy - on-prem Windows
- Microsoft Intune (MEM) - cloud-managed Windows, macOS, iOS, Android, Linux
- System Center Configuration Manager (SCCM) - hybrid
- Local script - pilot/manual
- VDI - non-persistent VMs with VDI onboarding
- Server via Defender for Cloud - automatic for Azure Arc / Defender for Servers
- macOS, Linux - shell scripts and config profiles
Attack Surface Reduction (ASR) Rules¶
15+ rules to block common attack vectors: - Block credential stealing from LSASS - Block Office apps from creating child processes - Block Office apps from creating executable content - Block executable content from email/webmail - Block JavaScript or VBScript from launching downloaded executable content - Block process creations from PsExec and WMI - Block persistence through WMI event subscription - Block untrusted/unsigned processes from USB - Use advanced protection against ransomware
Modes: Audit, Block, Warn (limited), Disabled. Pilot in Audit, then enforce.
Endpoint security policies (in Intune/MEM)¶
- Antivirus
- Disk encryption
- Firewall
- Endpoint detection and response (telemetry, isolation tag)
- Attack surface reduction
- Account protection
- Device compliance
Threat and Vulnerability Management (TVM)¶
Built-in risk-based vuln management: - Continuous discovery of CVEs across installed software - Exposure score (lower is better, opposite of secure score) - Recommendations prioritized by exploit availability, breach status, asset criticality - Browser extension assessment - Network share configuration assessment - Authenticated scan for Windows and network devices - Integration with Intune to push remediation
Live Response¶
Remote shell to investigate or remediate endpoints: - Browse file system - Get/put files - Run scripts (curated library) - Run binaries from approved library - View running processes, services, scheduled tasks, persistence locations - Logged in DeviceEvents table
Permissions: - Live response basic - browse, copy files, run commands - Live response advanced - run unsigned scripts, run executable
Web Content Filtering¶
Category-based blocking (Adult, Gambling, Liability, etc.) at network egress on managed endpoints. Reports surface in MDE > Reports > Web protection.
Device Control¶
USB removable storage policies, printer redirection control, Bluetooth control. Defined via Intune or local Group Policy.
Indicators (IOCs)¶
Block hashes, IPs, URLs, domains, certificates organization-wide. Limits: 15,000 indicators per type per tenant. Set via portal, MDE API, or via Defender XDR Threat Intelligence integration.
Defender for Cloud + Defender for Endpoint Integration¶
Defender for Servers Plan 1 includes MDE deployment to Azure VMs and Arc-enabled servers automatically. The MDE agent provides:
- EDR telemetry to Defender XDR (advanced hunting)
- Vulnerability assessment to Defender for Cloud
- Behavioral analytics to both surfaces
- Live response capability
- ASR rules
Plan 2 adds JIT VM access, adaptive application controls (machine learning allowlist), network hardening (suggested NSG tightening), and 500MB/day Sentinel ingestion of selected MDC tables.
Microsoft Defender for Identity (MDI)¶
Detects identity-based attacks against on-prem AD, AD FS, AD CS, and Entra Connect:
- Sensors installed on Domain Controllers, AD FS, AD CS servers (lightweight)
- Detects: DCSync, DCShadow, Golden Ticket, Silver Ticket, Pass-the-Hash, Pass-the-Ticket, Skeleton Key, AS-REP roasting, Kerberoasting, NTLM relay, suspicious LDAP/SAMR queries
- ITDR (Identity Threat Detection and Response) capability set
- Integration with Entra ID Protection for hybrid identity risk scoring
- Action accounts for response (disable user, force password reset on-prem)
- Honeytoken accounts to flag adversary recon
Microsoft Defender for Office 365 (MDO)¶
Email and collaboration security:
- Safe Links (URL rewriting and time-of-click protection)
- Safe Attachments (sandbox detonation)
- Anti-phishing policies (impersonation protection, spoof intelligence, mailbox intelligence)
- Anti-spam, anti-malware policies
- Threat Explorer / Real-time Detections for hunting and remediation
- Attack Simulation Training for end-user phishing exercises
- Quarantine management
P1: Safe Links, Safe Attachments, anti-phish basic. P2: Adds Threat Explorer, AIR, Attack Simulation, campaign views.
Microsoft Defender for Cloud Apps (MDA, formerly MCAS)¶
Cloud Access Security Broker (CASB):
- App discovery (analyzes proxy/firewall logs to find shadow IT)
- Cloud app catalog with risk scoring
- Connected apps via APIs (M365, Salesforce, Workday, GitHub, AWS, Azure, GCP) for monitoring and governance
- Conditional Access App Control for inline session monitoring (block download, block paste, etc.)
- Activity policies, file policies, anomaly policies, OAuth app policies
- App governance add-on for OAuth app risk management
Common Exam Pitfalls¶
- Confusing Defender for Cloud (workloads) with Defender for Cloud Apps (SaaS apps)
- Mixing MDI (on-prem AD) with Entra ID Protection (cloud identity)
- Selecting Servers P1 when JIT/adaptive controls (P2) are required
- Forgetting that Defender CSPM is a separate paid plan beyond foundational CSPM
- Not knowing that ASR rules have Audit mode for piloting
- Picking AV settings via SCCM when Intune endpoint security is the modern path
- Misidentifying the right Defender plan for a workload (Storage vs Cosmos vs SQL)
Quick Reference: Plan to Workload¶
| Workload | Plan |
|---|---|
| Azure VM, Arc server | Defender for Servers P1/P2 |
| Azure Storage | Defender for Storage |
| Azure SQL DB / Managed Instance / SQL on VMs | Defender for SQL |
| Cosmos DB | Defender for Cosmos DB |
| Postgres/MySQL/MariaDB | Defender for open-source DBs |
| AKS, ARO, EKS, GKE | Defender for Containers |
| Azure App Service | Defender for App Service |
| Azure Key Vault | Defender for Key Vault |
| Azure Resource Manager | Defender for Resource Manager |
| Azure DNS | Defender for DNS |
| Azure API Management | Defender for APIs |
| Azure OpenAI / AI Studio | Defender for AI |
| Endpoint (workstation/laptop) | MDE P1 or P2 (via M365 license) |
| Email/M365 collaboration | MDO P1 or P2 |
| On-prem AD | MDI |
| SaaS apps / shadow IT | MDA |