code
SC-500
level
Associate
pass
700 / 1000
cost
$165 USD
questions
~40–60
duration
100 min
renewal
free / annual
role
Cloud + AI Security Engineer
SC-500 has four domains and only one breaks 25% — storage, databases, and networking. That domain is also the one carried over most directly from AZ-500, so if you already hold AZ-500, your fastest path to a pass is: skim domains 1, 2, and the non-AI half of domain 3, then spend most of your study time on the AI-security sub-domain below.
This domain covers who can reach a resource and how you prove that access stays compliant over time. Expect scenarios describing an access requirement (temporary elevation, app-to-app trust, a secret that must never appear in plaintext) and asking which Entra ID or governance control satisfies it with the least standing privilege.
PIM
Privileged Identity Management — time-boxed, approval-gated activation of a privileged role instead of standing assignment. Reduces the window a role is usable, not just who can hold it.
Conditional access
Policy engine evaluating signals (location, device, risk) at sign-in time to allow, block, or require MFA — can target users, apps, or (new) Entra Agent ID.
Managed identity
Identity Azure manages for you on a resource — no credentials to rotate. Preferred over a service principal + secret for Azure-to-Azure calls.
Key Vault firewall
Network-level restriction on which VNets/IPs can reach the vault, independent of RBAC/access-policy permissions on the secrets themselves.
Defender for Key Vault
Threat detection on vault access patterns (e.g. unusual retrieval volume) — distinct from CSPM's static secret-scanning across resources.
⚠ decide: a scenario about reducing how long a role is usable is PIM; a scenario about whether a sign-in is allowed at all is conditional access.
The largest domain, and the one that maps most directly onto AZ-500's old “secure networking” and “secure compute, storage, and databases” domains. The recurring test is picking the right layer: does this scenario need a firewall rule, a private connection, or a detection plan?
Storage firewall
Restrict a storage account to specific VNets/IPs — prevention, configured on the resource itself.
Defender for Storage
Threat detection on a storage account (e.g. malware upload, anomalous access) — detection, not prevention.
Private endpoint
Brings a PaaS resource (Storage, SQL) into your VNet with a private IP — secures inbound access to that resource.
Private Link service
Exposes your own service privately to other VNets/tenants — the reverse direction of a private endpoint.
Entra Private Access
Secures user/hybrid access to private apps and on-prem resources — not a PaaS-to-PaaS or service-exposure control.
Network Watcher effective rules
Shows the actual merged NSG rules applying to a NIC — the way to verify what a scenario's NSG configuration really allows.
⚠ decide: ask “who is connecting to whom” — a resource reaching out privately needs a private endpoint; your service being reached needs Private Link; a user/hybrid client needs Entra Private Access.
This sub-domain has zero AZ-500 overlap. Every term here is new territory, even for AZ-500 veterans. The exam tests whether you know which control gates access to an agent, which one governs the data it touches, and which one detects after the fact.
Entra Agent ID
The identity type assigned to an AI agent. Conditional access can target it directly to restrict which users can invoke a given agent.
Purview DSPM
Data Security Posture Management — surfaces where Copilot/AI apps could expose over-shared SharePoint or other sensitive data before it becomes an incident.
Copilot Studio real-time protection
Runtime guardrails on a Copilot Studio agent's conversation, not a posture scan — catches unsafe output as it happens.
Foundry AI Gateway
Deployed via API Management in front of Microsoft Foundry — centralizes auth, rate limiting, and policy enforcement for model calls.
Foundry guardrails
Content filters and blocklists configured directly on a Foundry agent/model deployment.
Defender for AI Service
Cloud Workload Protection plan in Defender for Cloud — detects threats against AI services after deployment.
⚠ decide: who-can-invoke → Entra Agent ID + conditional access. what-data-is-exposed → Purview DSPM. detect-after-the-fact → Defender for AI Service / Defender XDR blast-radius analysis.
JIT VM access
Management ports stay closed by default; opened only for an approved, time-boxed window on request.
Azure Bastion
Browser-based RDP/SSH to a VM with no public IP exposed at all — different problem than JIT's time-boxing.
Agentless scanning
Defender for Servers scans a VM disk snapshot for vulnerabilities without installing an in-guest agent.
vTPM + secure boot
VM-level configuration blocking untrusted boot components and enabling measured/attested boot — a VM property, not a network control.
Azure Arc
Extends Defender for Servers and Azure Policy governance to on-prem/multicloud servers as if they were native Azure resources.
APIM back-end protection
Policies (auth, rate limits, IP restrictions) enforced at the API Management gateway in front of a back-end API or Function.
⚠ decide: “no public IP at all” is Bastion; “port open only for a short approved window” is JIT — they solve different threats and can be tested against each other.
The final domain checks whether you can find risk you already know about, discover risk you don't, and orchestrate a response once Sentinel flags something.
Defender CSPM
Posture/configuration risk across resources already in your inventory.
Defender Vulnerability Mgmt
Software vulnerability scanning specifically for Azure VMs.
Defender EASM
Discovers unmanaged/shadow assets outside your known inventory — the 'what don't we know about' surface.
Sentinel automation rule
Orchestration layer deciding which incidents trigger action and in what order; can tag/assign/close directly.
Sentinel playbook
The actual Logic App workflow an automation rule invokes to take a remediation action.
Security Copilot plugin
Extends Security Copilot's capabilities with additional data sources or actions inside a configured workspace.
⚠ decide: “assets we don't know exist” is EASM; “misconfiguration on assets we do know about” is CSPM; “a VM has an unpatched CVE” is Vulnerability Management.
$ test --readiness
Try 40 Free SC-500 Practice Questions
Scenario-based, calibrated harder than the real exam, including full AI-security coverage. No card required.
Start Free Practice →A short, practical list for the morning of the exam — not more material to learn, just the order to review it in when time is limited.
- Re-read the five “decide” callouts above — they cover the control-vs-control distinctions that decide close-call questions.
- Confirm you can explain, out loud, the difference between Entra Agent ID conditional access, Purview DSPM, and Defender for AI Service — the single highest-value comparison on the new AI-security sub-domain.
- Skim the acronym list once — recognizing DSPM, EASM, and JIT by sight saves reading time on scenario questions that assume you already know them.
- Do not cram new material the night before. SC-500 rewards calm scenario analysis over freshly memorized facts, same as AZ-500 did.
# Is this cheat sheet enough to pass SC-500 on its own?
No. SC-500 is scenario-heavy and assumes hands-on Azure security administration experience. Use this as a final-review consolidation, paired with hands-on time in a free Azure trial subscription and full-length practice questions — not as a substitute for either.
# What is the single biggest domain to prioritize?
Secure storage, databases, and networking (25–30%) — Storage/SQL security, Defender for Storage/Databases, NSGs/ASGs, private endpoints vs. Private Link vs. Entra Private Access, and Azure Firewall. It is the only domain above 25%, carried over almost entirely from AZ-500.
# What is the passing score and cost for SC-500?
The passing score is 700 out of 1000. The exam costs $165 USD, runs about 100 minutes, and has roughly 40–60 questions.
# What is genuinely new compared to AZ-500?
The full "Implement security for AI" sub-domain: Purview DSPM for Copilot risk, Copilot Studio real-time protection, Entra Agent ID conditional access and blast-radius analysis, Foundry AI Gateway in API Management, Defender for AI Service, Foundry guardrails, and the Data and AI security dashboard. Everything else on this cheat sheet maps closely onto what AZ-500 already tested.
# Is this cheat sheet current with the latest SC-500 exam objectives?
Yes. It is built directly from Microsoft's official SC-500 skills-measured outline, current as of July 2026 — the same source used in our full SC-500 study guide, so the domain weights and sub-skills here match what Microsoft Learn currently publishes.
# How does a cheat sheet differ from the full SC-500 study guide?
The study guide walks through each domain in depth with a week-by-week plan for building the underlying skills. This cheat sheet assumes you've already studied and is meant for rapid final review the week of your exam — short definitions and the specific distinctions that decide close-call questions, not step-by-step instructions.
# What should I review last, the night before the exam?
The five decision callouts under each domain below. They capture the control-vs-control distinctions (which service gates access, which one detects, which one governs data) that repeatedly separate correct answers from plausible-looking wrong ones — a higher-yield use of your last hour than re-reading full documentation.
about.md
MSCertQuiz provides 500 practice questions per certification, calibrated harder than the real exam so test day feels easier. Start with 40 free questions on SC-500 — no card required.