Prepare and test your skills

Prepare and test your skills

The provider's role in Google Cloud security and operations is to build, secure, and operate the trusted foundation that all customer workloads run on. Google Cloud (opens in a new tab) secures physical facilities, manages the core platform for each service model, and enforces the baseline controls that customers configure. Customer choices sit on top of this foundation, while Google Cloud keeps the foundation itself secure, available, and compliant.
Physical infrastructure security in Google Cloud means Google Cloud protects the data centers, hardware, networking equipment, and global infrastructure that host customer workloads. Google Cloud controls physical access to facilities, secures servers and network devices, and designs infrastructure with security as a primary requirement. Every virtual machine, container, database, and application depends on this same physical foundation, from Infrastructure as a Service (IaaS) to Software as a Service (SaaS). Customers never manage physical security in Google Cloud, so customer governance can focus on data, access, and configuration.
Foundational security controls in Google Cloud are the default protections and enforcement mechanisms that Google Cloud applies for every customer. Google Cloud applies default encryption to data, maintains infrastructure controls for how data is gathered, processed, and stored, and enforces the Identity and Access Management (IAM) policies that customers define. Enforcement flows from Google Cloud to customer workloads, so only identities the customer approves can administer workloads or access data. Customer assignment decides who gets which role, while Google Cloud enforcement makes that decision stick.
Google Cloud provides the following foundational responsibilities across all service models:
When a customer selects Assured Workloads controls for a compliance regime, Google Cloud enforces those storage and personnel-access restrictions for the protected data types in the configured resources. Organization policies flow from Google Cloud into the customer organization to constrain deployments, and access insights flow back to administrators to reveal risky grants.
Provider responsibility in Google Cloud grows as customers move from IaaS to Platform as a Service (PaaS) to SaaS, because Google Cloud manages more of the stack at each level. In IaaS, Google Cloud focuses on underlying infrastructure and physical security while the customer manages most operating system, application, and data controls. In PaaS, Google Cloud takes on the runtime environment plus additional platform controls, with application controls and IAM management shared. In SaaS, Google Cloud owns most security responsibilities for the application itself. This progression changes the tradeoff: higher-level services transfer operational work from the customer to Google Cloud.
| Service model | Example services | Google Cloud provider responsibility |
|---|---|---|
| Infrastructure as a Service (IaaS) | Compute Engine, Cloud Storage, Cloud VPN, Cloud Load Balancing, Cloud DNS | Secure underlying infrastructure and physical security; customer retains most security configuration |
| Platform as a Service (PaaS) | App Engine, Google Kubernetes Engine (GKE), BigQuery | Secure infrastructure plus runtime environment and more platform controls; application controls and IAM management are shared |
| Software as a Service (SaaS) | Google Workspace, Google Security Operations, Marketplace applications | Own most security responsibilities for the application; customer manages access and data placed in the application |
A lift-and-shift virtual machine on Compute Engine keeps patching and network configuration with the customer, while a BigQuery dataset or Google Workspace subscription places platform maintenance and availability with Google Cloud.
Exam tip: Google Cloud enforces the IAM policies you set, but you decide which identities get which roles; enforcement is the provider's job, assignment is the customer's job.
Trusted platform operations in Google Cloud mean Google Cloud builds, monitors, and maintains services so workloads stay secure and available. Google Cloud operates the global network, maintains service health, and limits internal access through strong authentication, infrastructure-wide controls, continuous monitoring, and audits by internal and independent third parties. Personnel access to customer data (opens in a new tab) is minimized and restricted to cases such as customer-initiated support, outage recovery, narrowly reviewed legal requests, and rare security or abuse reviews. During outages, engineers use logged tools to restore service, and supported services generate transparency logs that the customer can review.
Google Cloud also helps customers stay on a secure baseline through opinionated guidance and deployable controls. Secure blueprints provide infrastructure as code with security recommendations enabled by default, landing zone guides define resource hierarchy, identity onboarding, key management, and network structure, and services such as Security Command Center support threat detection. Guidance flows from Google Cloud blueprints into customer projects, and telemetry flows from customer resources back into monitoring and governance tools, so the platform and customer configuration improve together.
The cloud shared responsibility model divides security and management work between Google Cloud and the customer. The shared responsibility model (opens in a new tab) assigns physical infrastructure and the foundational platform to Google Cloud, while the customer always decides what data and workloads to protect and which controls to configure. As you move from on-premises infrastructure to IaaS, PaaS, and SaaS, more layers shift from the customer to Google Cloud.
The controls a customer must choose depend on three peer factors:
These factors determine how strict identity, data, and application controls must be in any deployment model.
On-premises infrastructure is the baseline where the customer controls and secures every layer. The customer owns the physical data center, hardware, networking, operating systems, applications, and data. Google Cloud has no operational role in this model, so all risk, patching, access control, and governance remain with the customer. This baseline shows the full amount of work that cloud models gradually transfer to the provider.
Infrastructure as a Service (IaaS) provides compute, storage, and network services on demand with pay-as-you-go pricing. Examples include Compute Engine for virtual machines, Cloud Storage for object storage, and networking services such as Cloud VPN, Cloud Load Balancing, and Cloud DNS. IaaS fits lift-and-shift migration of an existing on-premises workload or cases where you want specific virtual machines, databases, or network configurations.
In IaaS, the bulk of security responsibility remains with the customer, while Google Cloud is responsible for the underlying infrastructure and physical security. The customer still configures and secures the guest operating system, applications, data, identities, and virtual network settings. This model gives the customer the most control in the cloud, but also the most management work.
Platform as a Service (PaaS) provides the runtime environment where you can develop and run applications without managing the underlying infrastructure. Examples include App Engine, Google Kubernetes Engine (GKE), and BigQuery. PaaS fits building an application such as a website when you want to focus on development rather than servers and operating systems.
In PaaS, Google Cloud takes responsibility for more controls than in IaaS, including the hardware, host operating systems, and runtime platform. Responsibility for application-level controls and IAM is shared between Google Cloud and the customer and varies by service and features used. The customer remains responsible for data security and client protection, including what data is stored and who can access it.
Software as a Service (SaaS) provides ready-to-use online applications that you subscribe to or pay for in some way. Examples include Google Workspace, Google Security Operations, and third-party SaaS applications from Google Cloud Marketplace. SaaS fits when the enterprise needs to process workloads but does not have the expertise or business need to build the application itself.
In SaaS, Google Cloud owns the bulk of security responsibilities for the application, platform, and infrastructure. The boundary shifts so the customer no longer manages operating systems, runtimes, or application patching. The customer remains responsible for access controls and for the data the customer chooses to store in the application.
Exam tip: In IaaS you manage the operating system and applications, in PaaS you share application and IAM controls with Google Cloud, and in SaaS you manage only access and your data.
Control over hardware, networking, operating systems, applications, and data moves step by step from the customer to Google Cloud. The following comparison shows who manages each layer and how operational work decreases as provider control increases:
A layered stack diagram comparing responsibility across on-premises, IaaS, PaaS, and SaaS. Green layers are managed by Google Cloud, blue layers by the customer, and amber layers are shared. Provider responsibility grows from IaaS to SaaS, while the customer always keeps data and access responsibilities.
| Model | Google Cloud manages | Customer manages |
|---|---|---|
| Infrastructure as a Service (IaaS) | Physical security and underlying infrastructure | Operating systems, applications, data, IAM configuration, and virtual network settings |
| Platform as a Service (PaaS) | Infrastructure, operating systems, and runtime environment, plus shared application and IAM controls | Data security, client protection, and shared application and IAM configuration |
| Software as a Service (SaaS) | Infrastructure, platform, and application, including most security controls | Access controls and data stored in the application |
IaaS preserves control for custom configurations, PaaS trades some control for faster development, and SaaS trades the most control for the least management. Choosing a higher-level service therefore reduces customer maintenance but also reduces direct control over lower layers.
Customer responsibilities for configuration, governance, and data protection mean securing what the customer brings to the cloud while Google Cloud protects the cloud itself. The customer is always responsible for securing data, controlling who can access it, and correctly configuring applications and policies. Google Cloud secures the underlying infrastructure, but misconfigurations or governance gaps on the customer side can still cause unauthorized access, data loss, and other security incidents.
| Cloud model | Google Cloud protects | Customer protects |
|---|---|---|
| Infrastructure as a service (IaaS) | Hardware, storage, and network | Operating systems, applications, configurations, identity, and data above that foundation |
| Software as a service (SaaS) | Security of everything except data and its access and usage | Data and how data is accessed and used |
In IaaS the customer carries more of the stack, while in SaaS Google Cloud carries more of the stack. The customer responsibility for data does not disappear when Google Cloud manages more layers.
Data protection is the customer duty that never shifts to Google Cloud across on-premises, IaaS, PaaS, and SaaS. The customer owns its own data, including data received from users, and is responsible for privacy controls such as cookies, session management, and obtaining user permission. This duty includes deciding where data may reside, classifying confidential data by sensitivity, and locking down access so only approved identities and workloads can read or change sensitive data.
When data is misclassified or access is left too broad, the trust boundary between authorized users and everyone else breaks down. Data then flows from storage and applications to unintended destinations, which increases the risk of exposure and compliance failure. Strong classification and residency decisions let governance policies apply to the right datasets in the right locations.
Identity and Access Management (IAM) configuration is a core customer responsibility for users, groups, and service accounts. The customer must grant access according to the principle of least privilege, so each identity receives only the permissions needed for its task. For delivery and runtime operations, the customer should use dedicated, user-specified service accounts instead of default service accounts, and must safeguard any third-party integration tokens used by pipelines or applications.
In Google Kubernetes Engine (GKE) (opens in a new tab), this same duty extends to role-based access control (RBAC) and IAM policy for containers and pods. The customer maintains the RBAC and IAM policy that controls who can deploy, change, or view workloads, and rotates cluster credentials on schedule. If identity policy is too permissive or credentials are not rotated, an attacker who gains one identity can move laterally to workloads, data, and cluster operations.
Application and workload configuration means the customer secures source code, configuration files, build files, container images, and the running containers and pods. The customer must evaluate whether images meet security standards, use the latest supported image versions, and follow best practices for open source components and build configuration. In Cloud Deploy (opens in a new tab), the customer also ensures that replaced versions of embedded tools remain compatible with the deploy image and provides environmental details to Google Cloud when troubleshooting is requested.
The customer depends on verification controls to keep untrusted code from reaching production. That includes enabling and acting on vulnerability scanning for build artifacts, for example with Artifact Analysis, generating build provenance data, and enforcing deployment policies, for example with Binary Authorization, so only authorized and verified images are deployed. In GKE Standard, the customer keeps node pools enrolled in automatic upgrades and manually upgrades clusters and node pools when auto-upgrades are postponed by maintenance policies or when a patch is needed before it appears in the selected release channel.
Exam tip: The operating system underneath GKE containers is managed by Google Cloud, but the operating system inside customer containers and the application code itself remain customer responsibilities.
Governance is the customer process for identifying risks, meeting regulatory obligations, and continuously monitoring the environment. Before creating resources, the customer completes a risk assessment that catalogs organization-specific risks and selects needed security features, then repeats the analysis after deployment and whenever business needs, regulations, or threats change. Cloud-specific methods help here, such as the Cloud Controls Matrix (CCM) for risk assessment and Open Worldwide Application Security Project (OWASP) threat modeling for finding application vulnerabilities.
A typical compliance journey (opens in a new tab) follows an ordered sequence:
A three-step process flow showing the customer compliance journey: Assessment of regulatory obligations, Gap remediation using Assured Workloads, and Continual monitoring with tools like Google Cloud Observability and Security Command Center.
Governance gaps weaken visibility and control across environments, especially in hybrid and multicloud designs where each provider has different models and tooling. Without centralized monitoring and consistent policy, misconfigurations remain hidden in separate dashboards and are harder to prioritize. The customer must also keep Logging and Monitoring enabled on clusters, because if logs are unavailable, support for incidents is available on a best-effort basis only.
In Infrastructure as a Service (IaaS) the customer manages operating systems, applications, data, Identity and Access Management (IAM) configuration, and virtual network settings, in Platform as a Service (PaaS) application and Identity and Access Management (IAM) controls are shared with Google Cloud while the customer remains responsible for data security and client protection, and in Software as a Service (SaaS) the customer manages only access controls and data stored in the application. Google Cloud manages physical security and underlying infrastructure in Infrastructure as a Service (IaaS), adds operating systems and the runtime environment in Platform as a Service (PaaS), and owns most infrastructure, platform, and application security in Software as a Service (SaaS).
The customer decides which identities receive which roles under the principle of least privilege, while Google Cloud enforces the Identity and Access Management (IAM) policies the customer defines. Enforcement makes the customer assignment stick so only approved identities can administer workloads or access data.
The customer is always responsible for securing data, including data received from users, across on-premises, Infrastructure as a Service (IaaS), Platform as a Service (PaaS), and Software as a Service (SaaS). This duty includes deciding where data may reside, classifying confidential data by sensitivity, locking down access to approved identities and workloads, and managing privacy controls such as cookies, session management, and obtaining user permission.
A financial services company is migrating its customer portal to Google Cloud using Infrastructure as a Service (IaaS) and Platform as a Service (PaaS) solutions. During a pre-launch security review, the team identifies the following governance gaps:
Under the cloud shared responsibility model, which action is the customer's responsibility to remediate these risks and strengthen their security posture?
Last updated