Prepare and test your skills

Prepare and test your skills

Identity and Access Management (IAM) (opens in a new tab) and operational security practices decide who can do what in Google Cloud and how that secure state is kept over time. IAM and resource hierarchy design control access, organization policies and security perimeters enforce guardrails, and monitoring, logging, and threat detection find misconfigurations and attacks. Together they carry out the customer side of shared responsibility, where Google secures the underlying infrastructure and the customer secures data, applications, and access.
Controlling who can do what with IAM means granting only the access each person or workload needs through granular role-based access control. Access follows least privilege, so each identity receives only the permissions required for its job. Cloud Identity is a separate identity service that centrally manages users and groups and enables single sign-on to Google Cloud and thousands of third-party applications.
The IAM access model connects four parts, where a policy attached to a resource grants a role to a principal:
An entity-relationship diagram showing how an IAM allow policy binds a principal to a role on a specific resource, where the role contains permissions checked against the requested resource.
When a principal requests an action, Google Cloud checks the IAM allow policy on the requested resource. The request succeeds only if the policy binds that principal to a role containing the required permission, so the policy enforces the boundary between who is trusted and what they can change.
Exam tip: A role contains permissions, while an IAM allow policy binds principals to a role on a resource — the role defines what is allowed, the policy defines who gets it and where.
Organizing resources with hierarchy design means using the structure of Organization, folders, projects, and resources to set billing, visibility, and where controls apply. The Organization resource sits at the top and gives central visibility and control over everything in the organization. An organization contains folders, folders group related projects, and projects contain service resources such as Compute Engine virtual machines, Pub/Sub topics, Cloud Storage buckets, and Cloud VPN endpoints.
Hierarchy design reflects business processes and reporting structure, so teams can isolate workloads while keeping central control. Placing sensitive or regulated data with special access or residency needs inside dedicated Assured Workloads (opens in a new tab) folders or projects keeps it separate from general workloads and makes stricter controls easier to apply. Creating names and structures that do not expose personal data is itself a customer security duty.
Enforcing guardrails with policies and perimeters means setting limits that apply automatically across folders and projects. Organization policies let administrators configure constraints throughout the hierarchy, such as limiting where workloads can be deployed. Resource location constraints, for example, help ensure workloads run only in pre-approved regions to meet data residency needs. Policy Intelligence tools support enforcement by showing who has access to accounts and resources, so overly broad access can be found and fixed.
VPC Service Controls enforce a network and data boundary by creating a security perimeter around sensitive services and data. Traffic inside the perimeter can flow between protected services, while requests crossing the perimeter are denied unless explicitly allowed, which helps prevent data exfiltration and unauthorized access. Assured Workloads adds compliance-focused enforcement for regulated industries, including controls for where data is stored and which Google personnel can access data during legitimate work.
Monitoring operations and handling sensitive data means recording activity, detecting threats, and keeping regulated data in protected places. Security Command Center is the centralized service for continuous monitoring, threat detection, and response, combining threat intelligence, threat detection, and web scanning to find misconfigurations, vulnerabilities, and active threats. Because misconfiguration is a leading cause of breaches, continuous monitoring gives early warning before a weak setting becomes an incident.
Logging completes secure operations by making access visible for review and audit. Access Transparency logs show qualifying Google personnel access, including support, outage recovery, and other justified cases. Findings flow from Google Cloud resources to Security Command Center and logging systems, then back to administrators who adjust IAM policies, organization policies, or perimeters. Sensitive data such as account information, health information, personal identifiers, cardholder data, and ID numbers must stay inside compliant folders and must not be placed in resource names and IDs, project and folder names, descriptions and labels, log-based metrics, URIs and file paths, timestamps, user IDs, firewall rules, or scan configurations.
Core security controls and infrastructure protections are the layered defenses Google and the customer apply to physical infrastructure, virtual resources, identities, networks, and data. Google secures the underlying infrastructure, while customers configure access, place sensitive data correctly, and manage security services. Google's shared fate approach adds guidance, secure blueprints, and governance tools to help customers improve security outcomes.
Dividing work with shared responsibility and shared fate means splitting security tasks between Google Cloud and the customer and then working as partners over time. The customer understands regulatory and business requirements and decides which controls to configure based on compliance obligations, organizational standards and risk plans, and customer or vendor requirements. Google protects foundational infrastructure, applies default encryption and infrastructure controls, enforces the IAM policies the customer sets, and applies customer-selected Assured Workloads controls.
Customer responsibilities include:
Sensitive data can include account information, health information, personal identifiers, cardholder data, and ID numbers. Google protects many types, including Customer Data and Service Data, but workloads with elevated regulatory needs may require added tools.
Exam tip: In infrastructure as a service, the customer carries the bulk of security; in software as a service, Google owns the bulk, but the customer still controls access and the data stored in the application.
Protecting data with default encryption and infrastructure controls starts with protections Google applies without customer action. Google Cloud applies default encryption to data and provides strong protection for Customer Data and Service Data. Confidential Computing extends protection by encrypting data in use in memory, and customers with EU or sovereign needs can use custom encryption and key management options.
Google's infrastructure is built with security at its core and is fairly homogeneous, which lets Google use automation to apply controls and limit privileged access. Infrastructure-wide controls (opens in a new tab) include strong authentication for all access, cryptographic credentials for jobs, and application layer transport security (ALTS) for service-to-service requests. Google also uses pervasive control systems, detection and remediation of unauthorized access, continuous monitoring, violation alerting, and regular audits by internal and independent auditors.
Limiting movement with network and data boundaries means controlling where resources can run and where data can move. Google secures network security as part of the underlying infrastructure, and VPC Service Controls create security perimeters around sensitive services and data to prevent exfiltration. Resource location constraints help ensure workloads deploy only in pre-approved regions, while data residency rules control where data is stored and processed. Customers must account for regionality, including multiregional resources, when choosing a deployment target.
Assured Workloads and Sovereign Controls by Partners help place regulated data in compliant folders and add controls for locationally sensitive data. For EU configurations, added controls can include an EU data boundary, support routing to EU persons in EU locations, visibility into administrative access, policy-driven access approvals, and custom encryption options. Sensitive data should not be placed in fields that may fall outside technical controls, such as resource names, project names, labels, log metrics, VM sizes, URIs, timestamps, user IDs, firewall rules, or IAM policies.
Detecting threats and governing personnel access means monitoring for attacks while keeping Google's own access visible and controlled. Security Command Center Premium uses threat intelligence, threat detection, and web scanning to monitor resources and can resolve many threats quickly and automatically. Organization policies and Policy Intelligence provide governance across the hierarchy, while Access Transparency logs and Access Approval give visibility and control over Google personnel access to customer data.
Google personnel access can occur for customer-initiated support, outage or tool failure, third-party legal requests, and Google-initiated reviews. During support, Customer Care obtains only low-sensitive data, such as a bucket name if bucket access is lost. During outages, personnel can access data for backup or recovery, cannot bypass the allowlist, and their access is logged and audited. Legal requests are rare, only the legal team can justify them, and customers are notified when legally permitted. Together, default encryption, IAM, perimeters, monitoring, and audited recovery support resilience without removing oversight.
The shared responsibility model in Google Cloud is the division of security duties between Google and the customer: Google protects the cloud itself, and the customer secures what they bring to the cloud. The customer is always responsible for securing their data, while Google is always responsible for securing the underlying infrastructure. The model names which side owns each control so both parties know what they must configure, protect, and monitor.
Shifting security duties by service model means the line between Google's duties and customer duties moves with how much of the stack Google manages. In infrastructure as a service, only hardware, storage, and network belong to Google, so the customer owns more layers above them. In software as a service, Google owns almost everything except data and its access and usage, so customer duties shrink to protecting data and controlling who can reach it.
| Service model | Google Cloud is responsible for | Customer is responsible for |
|---|---|---|
| Infrastructure as a service (IaaS) | Hardware, storage, and network | Layers above those, including data and its access and usage |
| Platform as a service (PaaS) | More controls than in IaaS, including underlying platform | Shared application-level controls and IAM, plus data security and client protection |
| Software as a service (SaaS) | Security of everything except data and its access and usage | Data and its access and usage |
Examples include Compute Engine, Cloud Storage, Cloud VPN, Cloud Load Balancing, and Cloud DNS for IaaS; App Engine, Google Kubernetes Engine, and BigQuery for PaaS; and Google Workspace, Google Security Operations, and Marketplace applications for SaaS. A customer running workloads across multiple types must check which controls they still own for each one.
What Google secures versus what customers secure separates platform protection from workload protection. Google protects hardware, firmware, kernel, operating system, storage, and network, including physical data center security, default encryption at rest and in transit, and secure network components. Google also hardens and patches operating systems it provides, such as Container-Optimized OS, and maintains each service control plane, platform access controls, monitoring, and incident response.
The customer's side covers everything built, configured, and run on top of that infrastructure. Customer duties include complying with applicable laws, selecting appropriate images and verifying software beyond the base OS, using the latest base image versions, managing access and authentication, securing applications and source code with safe coding and vulnerability testing, monitoring their own applications, and configuring IAM with least privilege and dedicated service accounts. Google also restricts and logs its own administrative access for support using Access Transparency and Access Approval, and provides integrations such as IAM, Cloud Audit Logs, Observability, Cloud Key Management Service, and Security Command Center.
Making data protection and key choices means using Google's default encryption while the customer decides how keys and connections are managed. Google encrypts customer data at rest by default with no action required, and encrypts and authenticates data in transit at one or more network layers when it moves outside physical boundaries controlled by Google. All virtual machine-to-virtual machine traffic within a VPC network and between peered VPC networks is encrypted.
The customer chooses key management based on regulatory needs, hybrid environments, and data sensitivity:
Data is not encrypted by default while in use in memory, so organizations with confidential data can use Confidential Computing for a trusted execution environment. Traffic over Cloud Interconnect can be encrypted with MACsec, Cloud VPN traffic with IPsec, and application traffic with TLS and mutual TLS in Apigee or Cloud Service Mesh.
Specific services split duties the same way, with Google securing infrastructure and the customer securing use. Google Kubernetes Engine (GKE) (opens in a new tab) secures infrastructure, default encryption, node operating systems, the control plane, and Container Threat Detection, while the customer secures code, images, data, RBAC and IAM policy, containers and pods, credential rotation, and automatic upgrades. Cloud Build (opens in a new tab) secures isolated build environments and the control plane, while the customer secures source code, build files, tokens, least privilege, scanning, provenance, and deployment policies. Cloud Workstations (opens in a new tab) secures hardened default images and platform security, while the customer secures image choice, access controls, applications, and monitoring.
Exam tip: You are always responsible for securing your data, and Google is always responsible for securing the underlying infrastructure — this holds in every service model, including IaaS and SaaS.
Managing risk through shared fate partnership means going beyond a one-time split of duties to continuous collaboration. Cloud risk differs from on-premises risk because the provider carries hardware-stack risks the customer once carried, but boundaries still differ across IaaS, PaaS, and SaaS. Common challenges include misconfiguration as a direct cause of breaches, constantly changing products, overlapping controls across service types, changing regulations and acquisitions, unclear incident roles, and advanced persistent threats.
Shared fate treats security as an ongoing partnership with secure starting points and governance tools. It includes the enterprise foundations blueprint, secure blueprints deployed with infrastructure as code that have security recommendations enabled by default, the Google Cloud Well-Architected Framework, and landing zone guides for hierarchy, identity, key management, and network structure. The Risk Protection Program with Munich Re and Allianz includes Cyber Insurance Hub, which provides data-driven insights about security posture that customers can share with insurers for a quote. Governance products that support the model include Assured Workloads, Security Command Center Premium, Organization policies, Policy Intelligence, Confidential Computing, and Sovereign Controls by Partners.
An Identity and Access Management (IAM) role is a collection of permissions that defines what is allowed, while an Identity and Access Management (IAM) allow policy binds principals to a role on a specific resource and defines who gets it and where. When a principal requests an action, Google Cloud checks the allow policy on the requested resource and the request succeeds only if the policy binds that principal to a role containing the required permission.
Google Cloud is always responsible for securing the underlying infrastructure, while the customer is always responsible for securing their data and its access and usage. The division moves by service model: in infrastructure as a service Google is responsible for hardware, storage, and network, while in software as a service Google secures everything except data and its access and usage.
Virtual Private Cloud (VPC) Service Controls create a security perimeter around sensitive services and data to help prevent data exfiltration and unauthorized access. Traffic inside the perimeter can flow between protected services, while requests crossing the boundary are denied unless explicitly allowed.
Security Command Center is the centralized service for continuous monitoring, threat detection, and response in Google Cloud. It combines threat intelligence, threat detection, and web scanning to find misconfigurations, vulnerabilities, and active threats, giving early warning before a weak setting becomes an incident.
An enterprise is planning to deploy public-facing applications to Google Cloud. The security team requires that incoming internet traffic is automatically shielded from Denial-of-Service (DoS) attacks and that TLS encryption is terminated at the edge of the global network before requests are routed internally.
Which Google Cloud infrastructure component provides these foundational edge protections?
Customer-Managed Encryption Keys (CMEK)
VPC Service Controls
Security Command Center (SCC)
Google Front End (GFE)
Last updated