Cloud and AI Security Engineer (SC-500) Exam Questions

Domain 1: Manage identity, access, and governance (24%)

---

A security engineer configures Microsoft Entra Privileged Identity Management (PIM) so that a user is assigned as eligible for the Global Administrator role. The security team wants that whenever this user activates the role, they must supply a business justification, complete multifactor authentication, and receive approval from a designated approver before the role becomes active. Which PIM capability should the administrator configure to enforce this?

⬜ Access reviews
⬜ Role activation settings requiring justification, MFA, and approval
⬜ A permanent (non-expiring) eligible assignment
⬜ Conditional Access App Control

A financial services company wants to require that anyone eligible for an Azure administrator role uses only phishing-resistant authentication — such as a FIDO2 security key or certificate-based authentication — when signing in to the Azure portal, rather than accepting any generic MFA method like a phone call or SMS. Which Conditional Access control should be applied?

⬜ Require multifactor authentication
⬜ Require the device to be marked as compliant
⬜ Grant access requiring an authentication strength of “Phishing-resistant MFA”
⬜ A session control that limits sign-in frequency

A developer registers a single-tenant app in Microsoft Entra ID for a SaaS product, then reconfigures it as a multi-tenant application registration so customers in other tenants can use it. When an administrator in a customer’s tenant consents to the app for the first time, what object is created in that customer’s own Microsoft Entra tenant to represent the app and hold its granted permissions?

⬜ A duplicate copy of the app registration
⬜ A service principal (shown as an enterprise application) in the customer’s tenant
⬜ A new Conditional Access policy scoped to the app
⬜ A managed identity scoped to the customer’s subscription

Security monitoring discovers that a newly registered third-party application requested high-privilege Microsoft Graph application permissions, including Mail.Read and Files.ReadWrite.All, and that several regular (non-admin) employees were able to consent to it individually for the whole organization. Which configuration change would have prevented an individual, non-admin user from granting consent to an application requesting such high-privilege permissions?

⬜ Enabling the admin consent workflow so users can request administrator approval
⬜ Configuring user consent settings to block user consent, or to require admin consent for permissions classified as requiring admin approval
⬜ Turning on security defaults for all users
⬜ Creating an access review for the application’s assignments

A team runs 20 Azure Automation accounts, and every runbook across all of them needs to read secrets from the same Azure Key Vault. The team wants one identity, created a single time and granted the Key Vault role once, that can then be attached to all 20 Automation accounts — and deleting any one Automation account should not remove that identity from the rest. Which identity type meets this requirement?

⬜ A system-assigned managed identity configured on each Automation account
⬜ A single user-assigned managed identity, created once and assigned to every Automation account
⬜ A service principal authenticated with a client secret
⬜ Certificate-based authentication configured separately on each Automation account

An organization’s Azure Key Vault still uses the legacy vault access policy model, which only grants “all or nothing” permissions per principal across each object type (keys, secrets, certificates). The security team wants to grant one application read-only access to just secrets, layer Azure RBAC conditions on top later if needed, and ensure that every access grant change appears in the Azure Activity Log for auditing. Which access control model should the vault be switched to?

⬜ Keep vault access policies, but reduce the listed permissions
⬜ The Azure role-based access control (RBAC) permission model for Key Vault, using built-in roles such as Key Vault Secrets User
⬜ Storage account shared access signatures (SAS)
⬜ Microsoft Entra Conditional Access policies applied to the vault’s resource ID

A company enables the Azure Key Vault firewall, sets the default action to Deny, and adds only its office’s outbound IP address as an allowed rule. Afterward, Azure Backup and Azure Resource Manager deployments that read secrets from the vault start failing, even though those operations originate from Microsoft-managed infrastructure rather than the allowed office IP. What should the administrator configure so that specific trusted Microsoft services can bypass the firewall’s IP and virtual network rules, without exposing the vault to the public internet?

⬜ Add 0.0.0.0/0 as an allowed IP range
⬜ Enable “Allow trusted Microsoft services to bypass this firewall”
⬜ Disable the Key Vault firewall entirely
⬜ Create a private endpoint and delete every firewall rule

Microsoft Defender for Key Vault raises an alert showing that a service principal performed a “list secrets” operation on a production key vault and then immediately issued “get secret” calls against nearly every secret returned, a pattern that doesn’t match that principal’s historical usage. What kind of activity does this pattern typically indicate?

⬜ Secret dumping / credential harvesting following compromise of the service principal’s credentials
⬜ A brute-force password attack against the vault’s access policies
⬜ A routine Key Vault backup operation
⬜ A distributed denial-of-service (DDoS) attack against the vault’s public endpoint

A governance team wants every new Azure SQL Database resource created in a subscription to automatically have an auditing configuration deployed that sends logs to a central Log Analytics workspace — but resource creation itself must never be blocked, even if auditing hasn’t been configured yet. Which Azure Policy effect should the definition use?

⬜ Deny
⬜ Audit
⬜ DeployIfNotExists, using a managed identity granted to the policy assignment
⬜ Disabled

An administrator applies a ReadOnly lock to a production storage account to prevent accidental changes. Afterward, the operations team — who hold the Owner role on the subscription — report that they can no longer list the storage account’s access keys through the portal or the management API. What explains this behavior?

⬜ The ReadOnly lock revoked their Owner role assignment
⬜ A ReadOnly lock restricts every authorized user to Reader-equivalent permissions, and the List Keys operation is a POST call that a ReadOnly lock blocks
⬜ ReadOnly locks only affect delete operations, so this must be unrelated to the lock
⬜ The storage account’s network firewall is blocking their request

A company needs a support team to be able to create and manage app registrations and update Conditional Access policies inside Microsoft Entra ID, but that team must have no permissions on any Azure subscription or resource group. What should be created and assigned to meet this requirement?

⬜ A custom Azure role assigned at the management group scope
⬜ A custom Microsoft Entra role, built from allowed Microsoft Graph directory permissions and assigned at the Microsoft Entra organization (tenant) scope
⬜ The built-in Owner role scoped to a single resource group
⬜ A Key Vault access policy

As part of a ransomware-readiness review, a company wants to ensure that even if a Backup Administrator’s account is compromised, that account alone cannot disable soft delete or delete backup data from a Recovery Services vault — a separate, independent approver must authorize such critical operations first. Which Azure Backup security feature should be implemented, using a Resource Guard deployed in a different subscription or tenant from the vault?

⬜ Cross Region Restore
⬜ Multi-user authorization (MUA)
⬜ Immutable storage for blobs
⬜ Backup encryption using customer-managed keys

Domain 2: Secure storage, databases, and networking (28%)

---

A blob container in an Azure Storage account currently grants third-party partners access through several service SAS tokens, each valid for 12 months. The security team wants a way to instantly revoke all of those tokens before their expiry if a partner relationship ends, without regenerating the storage account’s access keys (which would break every other application using them). What should they configure?

⬜ A new account SAS with a shorter expiry
⬜ A stored access policy on the container, associated with the existing service SAS tokens
⬜ Azure Storage firewall IP rules
⬜ A private endpoint for the storage account

An administrator configures network rules on a storage account, setting the default action to Deny and allowing only a specific corporate IP range. After this change, all traffic that isn’t from an explicitly allowed source is denied access to the account’s public endpoint. Which statement correctly describes the previous, default behavior of the storage account’s network access before any rules were configured?

⬜ By default, before any network rules are configured, the storage account accepts connections from any network
⬜ By default, storage accounts deny all public network traffic until explicit IP rules are added
⬜ By default, only traffic from the same Azure region as the storage account is allowed
⬜ By default, only Azure Virtual Machines in a peered virtual network can reach the storage account

A company enables Microsoft Defender for Storage on a storage account used for user file uploads in a web application. They want any newly uploaded blob to be automatically scanned for malware the moment it lands, with the result of each scan written back to the blob so downstream logic can decide whether to allow the file to be served. Which Defender for Storage capability provides this, and how are results most simply surfaced per blob?

⬜ On-upload malware scanning, with results written as blob index tags on each scanned blob
⬜ On-demand malware scanning, triggered manually once a week against the whole account
⬜ Defender for Storage activity monitoring alerts only, with no per-file scan result
⬜ Azure Policy guest configuration extensions installed on the storage account

A DBA reviews a newly provisioned Azure SQL Database and needs to confirm whether data at rest — including the database files, transaction log, and automated backups — is encrypted, and if so, by which feature and at what level it’s configured. What should the DBA expect?

⬜ Transparent Data Encryption (TDE) is enabled by default on the new database, with the TDE protector managed at the logical server level and inherited by all databases on that server
⬜ Encryption at rest must be manually enabled using Always Encrypted before any data is protected
⬜ Dynamic Data Masking automatically encrypts the underlying data files at rest
⬜ No encryption at rest is applied unless a customer-managed key is explicitly configured first

An Azure SQL Database has server-level auditing enabled, writing all audit logs to a Log Analytics workspace. An administrator additionally enables database-level auditing on one specific database, pointing it at a different storage account, intending it to “override” the server policy for that database. What actually happens?

⬜ The database-level policy replaces the server-level policy for that database, and only the database-level logs are written
⬜ Both policies apply independently and at the same time, so the database is audited twice — once under each policy, writing to both destinations
⬜ Enabling database-level auditing automatically disables server-level auditing tenant-wide
⬜ The two policies conflict and Azure blocks the database-level policy from being enabled

A security team enables Microsoft Defender for Databases on an Azure SQL server and wants an automated, recurring assessment that identifies database misconfigurations and vulnerabilities — such as excessive permissions or missing auditing — and provides scripts to remediate each finding. Which capability provides this?

⬜ SQL vulnerability assessment
⬜ Dynamic Data Masking
⬜ Azure SQL auditing alone
⬜ Microsoft Entra Conditional Access for the SQL server’s sign-in endpoint

A network team manages several web-tier and database-tier virtual machines within the same subnet, which currently share one large NSG with rules referencing individual VM IP addresses. As VMs are added and removed frequently, the team wants to write and maintain rules based on each VM’s application role instead of its ever-changing IP address. Which Azure networking feature should they use inside their NSG rules to achieve this?

⬜ Azure Virtual Network Manager network groups
⬜ Application security groups (ASGs) as the source/destination in NSG rules
⬜ Azure Firewall DNAT rules
⬜ Azure Private Link

A central network security team uses Azure Virtual Network Manager to deploy a security admin rule with the action set to “Always Allow” for a specific network group of virtual networks, intended to guarantee that health-monitoring traffic is never blocked anywhere in scope. Individual application teams still manage their own NSGs on subnets within those virtual networks. If an application team’s NSG has an explicit Deny rule that would otherwise block that same traffic, what happens?

⬜ The traffic is still allowed — “Always Allow” security admin rules take effect before NSGs and cannot be overridden by an NSG’s Deny rule
⬜ The NSG’s Deny rule wins because NSGs are always evaluated first
⬜ The two conflicting rules cause the connection to be dropped as unresolved
⬜ Security admin rules only take effect if no NSG exists on the subnet

A company is designing a hub-and-spoke network built on Azure Virtual WAN and wants all internet-bound and inter-spoke traffic routed through a single, centrally managed Azure Firewall instance placed in the Virtual WAN hub itself, rather than deploying a separate firewall VNet alongside the hub. Which Azure Firewall Manager concept describes deploying Azure Firewall directly inside a Virtual WAN hub this way?

⬜ A secured virtual hub
⬜ A secured hub virtual network (non-Virtual WAN)
⬜ Azure Firewall Premium TLS inspection
⬜ Network security group flow logs

A company’s on-premises VPN device only supports a narrow, non-default combination of IKE encryption, integrity, and Diffie-Hellman group settings mandated by their security policy, which doesn’t match any of the default policy combinations Azure proposes during IKE negotiation for a site-to-site VPN connection. What should be configured on the Azure VPN Gateway connection to make the negotiation succeed?

⬜ A custom IPsec/IKE policy on the connection, matching the exact algorithms and parameters required
⬜ A higher VPN Gateway SKU with no other changes
⬜ Azure Firewall threat intelligence in Alert and Deny mode
⬜ An NSG rule allowing UDP port 500 and 4500

An organization wants remote employees to securely reach a small set of specific internal line-of-business applications, identified by FQDN and IP ranges, without giving them a full network-level tunnel into the corporate network the way a traditional VPN would, and while still being able to apply Conditional Access policies to that access. Which solution, and which required client component, fits this scenario?

⬜ Microsoft Entra Private Access using Quick Access, with the Global Secure Access Client installed on the device
⬜ A traditional point-to-site VPN gateway with a full split-tunnel configuration
⬜ Azure Bastion published to the public internet
⬜ A Private Link service exposed to the employees’ home IP addresses

A company hosts an Azure SQL Database and wants to eliminate all public internet exposure to it, so that traffic from an application in a peered virtual network reaches the database using a private IP address drawn from that same virtual network’s address space, appearing as if the database resource were located directly inside the VNet. Which technology should be configured for the Azure SQL Database?

⬜ An Azure private endpoint
⬜ A service endpoint for Microsoft.Sql
⬜ A public IP-based firewall rule scoped to the peered VNet’s NAT gateway address
⬜ Network Watcher effective security rules

A software vendor runs its own multi-tenant service behind a Standard Load Balancer in its own virtual network and wants to let many separate customer organizations connect to that service privately, without the vendor’s service being exposed to the public internet and without requiring VNet peering with each customer. From the provider’s side, which Azure networking capability should the vendor configure to publish its service for private consumption by many customer virtual networks?

⬜ An Azure Private Link service
⬜ An Azure private endpoint
⬜ A virtual network service endpoint
⬜ Azure Virtual Network Manager connectivity configuration

An Azure Firewall administrator enables threat intelligence-based filtering on a firewall without changing any other setting, expecting malicious traffic to and from known-bad IP addresses, domains, and URLs to be automatically blocked. After testing, they discover that matching traffic is being logged but not actually blocked. What is the correct explanation?

⬜ Threat intelligence-based filtering defaults to “Alert only” mode, so it must be explicitly changed to “Alert and Deny” to block matching traffic
⬜ Threat intelligence-based filtering can only ever alert; blocking must be done through a separate DNAT rule
⬜ Threat intelligence-based filtering requires Azure Firewall Premium, and the administrator is using the Basic SKU
⬜ Alerts are a display-only feature and blocking happens automatically regardless of mode

A troubleshooting engineer needs to determine exactly why traffic to a specific virtual machine’s network interface is being blocked, given that the VM’s NIC has its own NSG, the subnet it sits in has a separate NSG, and the virtual network is also in scope for an Azure Virtual Network Manager security admin configuration. Rather than manually cross-referencing all three rule sets, which single Azure Network Watcher capability should the engineer use to see the combined, effective set of rules actually being applied to that network interface?

⬜ Effective security rules
⬜ IP flow verify
⬜ Connection monitor
⬜ NSG flow logs traffic analytics

Domain 3: Secure compute (24%)

---

Contoso has rolled out Microsoft 365 Copilot to its entire workforce. The security team is concerned that Copilot could summarize or surface sensitive files from SharePoint sites where permissions are overly broad (for example, sites shared with “Everyone except external users”), exposing data to employees who should not see it. The team wants a Microsoft solution that specifically scans SharePoint content and permissions to identify this kind of AI-related data overexposure before it becomes a Copilot risk.

⬜ Microsoft Entra Permissions Management
⬜ Azure Policy compliance dashboard
⬜ Microsoft Purview Data Security Posture Management (DSPM) for AI
⬜ Microsoft Defender for Cloud Apps app governance

A company enabled Microsoft Copilot Studio agents that are published to employees and can call external actions. The SOC wants a capability that inspects each conversation turn while the agent is running in production and can automatically block or flag a session if the agent starts leaking sensitive data or is being jailbroken, rather than only reviewing logs after the fact.

⬜ Microsoft Purview Communication Compliance
⬜ Real-time protection for agents in Microsoft Defender for Cloud Apps
⬜ Copilot Studio analytics dashboard
⬜ Microsoft Sentinel scheduled analytics rules

An organization has deployed several autonomous AI agents that authenticate using their own Microsoft Entra Agent ID identities via the client credentials flow (no signed-in user is present). The identity team wants to enforce Conditional Access so that a high-risk agent identity can be blocked from obtaining new access tokens. A junior admin proposes simply adding the agent identities to the existing “All users” Conditional Access policy that covers human accounts.

⬜ This works as designed, because “All users” policies automatically include every agent identity and agent user account in the tenant
⬜ Conditional Access cannot evaluate any signal for agent identities, so no policy can block them
⬜ A separate policy must target agent identities directly, because policies scoped to “All users” do not automatically include agent identities or agents’ user accounts
⬜ Enabling security defaults is required first, since security defaults extend all existing Conditional Access policies to cover agent identities

A SOC analyst is triaging an incident suggesting that an AI agent’s Microsoft Entra Agent ID may have been compromised through a prompt-injection attack. Before taking remediation action, the analyst needs to understand the full scope of impact — what other entities, resources, and downstream systems the agent could have touched or reached as a result of the compromise.

⬜ Run a Kusto query against the SigninLogs table in Log Analytics
⬜ Use the incident graph and investigation experience for AI agent threats in Microsoft Defender XDR
⬜ Review the Copilot Studio agent’s conversation transcript export
⬜ Open the Azure Advisor recommendations blade for the resource group

Several internal applications call a shared Microsoft Foundry model deployment through a single Azure API Management instance acting as an AI gateway. Recently, one high-volume application has been consuming so many tokens per minute that other applications sharing the same backend are getting throttled by the model provider. The platform team wants to cap token consumption per application at the gateway layer itself, before requests reach the backend.

⬜ Configure the rate-limit-by-key policy based on request count
⬜ Configure the llm-token-limit policy scoped to each subscription key
⬜ Enable Azure Front Door WAF rate limiting rules
⬜ Increase the Foundry deployment’s provisioned throughput units (PTUs)

A company built a customer-facing chatbot on Azure OpenAI. After enabling Microsoft Defender for Cloud’s threat protection for AI workloads on the resource, the security team starts receiving alerts whenever a user’s prompt appears to be attempting to bypass the model’s safety guidelines through cleverly crafted invisible characters or by trying to extract the system prompt. Which capability is generating these alerts?

⬜ Microsoft Defender for AI Services, using signals such as Azure AI Content Safety Prompt Shields
⬜ Microsoft Purview Insider Risk Management
⬜ Azure DDoS Protection Standard
⬜ Microsoft Entra ID Protection risk detections

A finance team requires that data written to a VM’s temporary disk and the disk caches used by OS and data disks be encrypted at the Azure Storage infrastructure level, transparently, for any guest operating system, without installing an in-guest extension and without needing the VM to be running. Which encryption capability meets this requirement, as opposed to Azure Disk Encryption (which uses BitLocker/DM-Crypt inside the guest OS)?

⬜ Azure Disk Encryption with a customer-managed key
⬜ Encryption at host
⬜ Storage Service Encryption with platform-managed keys only
⬜ Double encryption at rest for managed disks

A security architect wants to reduce the attack surface of a fleet of Azure VMs that currently have RDP (3389) and SSH (22) open to specific management IP ranges around the clock. The requirement is to keep the management ports closed by default in the network security group and only open them for a specific source IP and a limited time window when an authorized administrator explicitly requests access through Microsoft Defender for Cloud.

⬜ Azure Bastion in standard SKU
⬜ Just-in-time (JIT) VM access
⬜ Azure Firewall with a DNAT rule
⬜ Network Security Group flow logs

An organization enables Defender for Servers Plan 2 and turns on agentless machine scanning across its Azure and multicloud VM estate. A team member is documenting how this capability actually works so colleagues understand what to expect. (Choose 2.)

⬜ It works by taking a snapshot of the VM’s managed disks and analyzing it out-of-band, so it has no performance impact on the running workload
⬜ It requires the legacy Log Analytics (MMA) agent to be installed inside every guest operating system
⬜ It only works for Azure virtual machines and cannot assess AWS EC2 or GCP Compute Engine instances
⬜ It can surface vulnerabilities and exposed secrets on virtual machines even while they are powered off, since it doesn’t rely on an in-guest agent

A platform team has onboarded an AKS cluster to Microsoft Defender for Containers. Azure Policy add-on checks are already flagging misconfigured pod security settings at deployment time, but the security team also wants to detect suspicious behavior while containers are actually running, such as a process spawning a cryptomining binary or an unexpected exec into a container. Which capability specifically addresses this runtime detection need?

⬜ Azure Container Registry vulnerability scanning
⬜ The Defender sensor (Defender agent) deployed to the AKS cluster for runtime threat detection
⬜ Azure Policy’s built-in Kubernetes misconfiguration definitions
⬜ Azure Monitor Container insights performance metrics

A developer wants every request to an HTTP-triggered Azure Function to be authenticated against Microsoft Entra ID before the function’s code executes, without writing any authentication logic inside the function itself and without introducing an API gateway in front of it. Which built-in capability of the Function App should be configured?

⬜ Function-level keys (function or host keys) on the HTTP trigger
⬜ App Service authentication (“Easy Auth”) on the Function App
⬜ A managed identity assigned to the Function App
⬜ Virtual network integration with a private endpoint

A public-facing web application sits behind Azure Application Gateway with Web Application Firewall (WAF) enabled using the OWASP managed rule set. During a recent penetration test, SQL injection payloads that matched WAF rules were logged but were still passed through to the application. The team wants matching malicious requests to actually be blocked going forward.

⬜ Add the application’s public IP to the WAF’s exclusion list
⬜ Increase the WAF rule set’s anomaly scoring threshold
⬜ Change the WAF policy mode from Detection to Prevention
⬜ Disable the OWASP rule set and rely only on custom rules

Domain 4: Manage and monitor security posture (24%)

---

A cloud security team has enabled the Defender CSPM plan across its Azure subscriptions. They notice Defender for Cloud has raised a single high-severity finding that combines three individually lower-risk conditions: a VM exposed directly to the internet, a known exploitable vulnerability on that VM, and a system-assigned managed identity on the VM with Owner rights over a Key Vault holding production secrets. Which Defender CSPM capability produced this combined, contextualized finding?

⬜ Secure score recommendations
⬜ Attack path analysis
⬜ Regulatory compliance dashboard
⬜ Workload protection alerts

An organization must demonstrate to an external auditor that its Azure environment meets the specific controls defined by PCI DSS 4.0, showing a pass/fail status per control rather than a generic hardening score. The compliance team also needs to track a custom internal security benchmark that isn’t included among Defender for Cloud’s built-in standards.

⬜ Rely solely on the Secure Score percentage shown on the Defender for Cloud overview page
⬜ Use the regulatory compliance dashboard, and assign a custom standard built from a custom Azure Policy initiative for the internal benchmark
⬜ Export Azure Advisor recommendations to Excel and manually map them to PCI DSS controls
⬜ Enable Microsoft Purview Compliance Manager exclusively, since Defender for Cloud cannot track compliance standards

A company runs dozens of AWS accounts under a single AWS Organizations management account. Rather than configuring a separate connector for every member account, the cloud security team wants Defender for Cloud to automatically discover and protect all current and future member accounts in the organization through one onboarding process.

⬜ Create a “Single account” connector once for each of the AWS accounts individually
⬜ Create the native AWS connector using the “Management account” option, which onboards the AWS Organization via a CloudFormation StackSet
⬜ Install the Log Analytics agent manually on every EC2 instance across all accounts
⬜ Use Azure Lighthouse to delegate access to each AWS account

A CISO is concerned that various business units may have stood up internet-facing web applications, domains, and certificates without informing central IT, creating unmonitored “shadow IT” exposure. The CISO wants a solution that discovers this externally visible footprint the way an outside attacker would see it — without needing existing inventory data or credentials to those resources.

⬜ Microsoft Defender for Cloud’s asset inventory blade
⬜ Microsoft Defender External Attack Surface Management (EASM)
⬜ Azure Resource Graph queries against subscriptions
⬜ Microsoft Purview Data Map

A SOC lead is onboarding new tier-1 analysts to Microsoft Sentinel. The analysts need to be able to view incidents, workbooks, and analytics rules, and to triage work by assigning, tagging, and closing incidents. They should not be able to create or edit analytics rules, install Content hub solutions, or manage other Sentinel resources. Which single built-in Azure role should be assigned to meet exactly this requirement?

⬜ Microsoft Sentinel Reader
⬜ Microsoft Sentinel Responder
⬜ Microsoft Sentinel Contributor
⬜ Microsoft Sentinel Playbook Operator

A Sentinel engineer wants to quickly deploy a pre-packaged bundle for a specific security product that includes a data connector, matching analytics rule templates, a workbook, and a playbook template, instead of manually building the data connector, writing analytics rules from scratch, and creating a workbook individually.

⬜ Manually author a custom Log Analytics data collection rule
⬜ Deploy the vendor’s solution from the Microsoft Sentinel Content hub
⬜ Import a KQL query pack from GitHub into the Logs blade
⬜ Create a new Azure Monitor workbook template from scratch

An organization runs an on-premises Linux-based firewall appliance that emits logs in Common Event Format (CEF) over syslog. The Sentinel engineer needs to ingest these logs into Microsoft Sentinel and has been told to avoid the legacy Log Analytics (MMA/OMS) agent, which is being retired, in favor of the current supported collection method.

⬜ Point the firewall directly at the Log Analytics workspace’s ingestion endpoint using the HTTP Data Collector API
⬜ Deploy the Azure Monitor Agent (AMA) on a Linux log forwarder and configure the CEF/Syslog via AMA data connector with a data collection rule
⬜ Install Microsoft Sentinel’s UEBA sensor on the firewall appliance itself
⬜ Configure Microsoft Defender for Cloud Apps to proxy the firewall’s log traffic

A company has hundreds of on-premises Windows servers that are not connected to Azure Arc and cannot each run the Azure Monitor Agent directly due to network segmentation. They already operate a central Windows Event Collector (WEC) server that aggregates security events from all of these servers using Windows Event Forwarding (WEF). The Sentinel engineer needs to get these forwarded events into Sentinel’s Windows Security Events table without deploying an agent to every source server.

⬜ Enable the deprecated legacy “Security Events” connector on each of the hundreds of source servers
⬜ Install the Azure Monitor Agent on the WEC/WEF collector server and configure the Windows Security Events via AMA connector with a data collection rule scoped to that collector
⬜ Manually export .evtx files from each server and upload them to a storage account
⬜ Enable Microsoft Defender for Endpoint on the WEC server only and rely on its telemetry as a substitute

A Sentinel engineer is designing incident response automation and needs to decide what belongs in an automation rule versus a playbook. Which two statements correctly describe the capabilities of automation rules? (Choose 2.)

⬜ An automation rule can automatically tag, assign, suppress, or close incidents based on conditions, without any code
⬜ An automation rule can orchestrate the order in which multiple playbooks run in response to an incident
⬜ An automation rule can call external, non-Microsoft REST APIs to remediate a resource
⬜ An automation rule replaces the need for Azure Logic Apps entirely, since all remediation logic runs natively inside the rule

An organization’s Security Copilot workspace currently has two people assigned the Copilot Owner role. An admin wants to remove one of them so that only a single named owner remains, believing this simplifies governance.

⬜ This is fine; Security Copilot has no minimum owner requirement
⬜ This isn’t possible as described; Security Copilot enforces a minimum of two owners at all times to prevent the workspace from being left without administrative access
⬜ This is fine, but only if the remaining owner also holds the Microsoft Entra Global Administrator role
⬜ This is fine, but the removal must be performed from the Microsoft Entra admin center rather than Security Copilot settings

A Security Copilot Contributor tries to run a prompt using the Microsoft Sentinel plugin to pull incident data, but receives no results even though the plugin is enabled for the tenant. Investigation shows the user has no Azure RBAC role assigned on the relevant Sentinel-enabled workspace. What is the most likely explanation, based on how Security Copilot plugin access works?

⬜ Security Copilot plugins bypass the underlying product’s permission model entirely, so this shouldn’t happen
⬜ Security Copilot uses on-behalf-of authentication for plugins, so the signed-in user still needs the underlying product’s own permissions (such as Microsoft Sentinel Reader) to retrieve that data
⬜ Only Copilot Owners can ever use any plugin, regardless of underlying product permissions
⬜ Plugins only work for users licensed with Microsoft 365 E5, independent of Azure RBAC

A SOC manager wants to add a ready-made, prebuilt Security Copilot agent — such as a phishing triage agent — without building custom promptbooks or logic from scratch, and wants to browse a catalog of such Microsoft and partner-built agents before deploying one into the tenant’s workspace.

⬜ Write a new custom plugin manifest and upload it manually
⬜ Browse and deploy a prebuilt agent from the Microsoft Security Store
⬜ Duplicate an existing promptbook and rename it
⬜ Request the agent be built through Azure Logic Apps custom connectors

You restrict the Microsoft Defender XDR plugin to owners only in your Security Copilot workspace. Which statement best describes the result?

⬜ The plugin is disabled only for new contributors added after the restriction
⬜ Analysts using Security Copilot in the Microsoft Defender portal see a restricted or degraded embedded experience
⬜ The plugin is removed from all workspaces in the tenant, not just this one
⬜ Existing contributors must individually reconfigure their own plugin settings to regain access

You need to assign the SOC team lead permissions to configure plugins and manage user access within the SOC workspace. Which Security Copilot role provides these capabilities?

⬜ Contributor
⬜ Security Administrator
⬜ Owner
⬜ Azure Contributor