Prepare and test your skills

Prepare and test your skills

Google’s defense-in-depth model is a security strategy that uses multiple independent layers so that if one control fails, others still protect data and systems. In Google Cloud (opens in a new tab), these layers span physical data centers and custom hardware, the global network, data encryption, identity and access, and security operations. Google secures the underlying cloud infrastructure, while customers secure what they put in the cloud, including their data and how it is accessed and used.
The hardware and physical infrastructure layer is the foundation that protects data centers, servers, storage devices, and custom chips that run Google Cloud. Physical security is layered with perimeter controls such as fencing, vehicle barriers, metal detectors, alarms, and electronic access cards, plus detection controls such as laser beam intrusion detection and 24/7 high-resolution cameras. Human controls include trained guards who pass background checks, and entry to the data center floor requires a security corridor with multi-factor access using badges and biometrics, open only to approved employees with specific roles.
Inside the data center, Google reduces vulnerabilities through custom server design and strict hardware controls. Servers omit unnecessary components such as video cards, chipsets, and peripheral connectors, and component vendors are vetted and audited. Titan security chips (opens in a new tab) help identify and authenticate legitimate Google devices at the hardware level, verify boot code, and provide secure key storage, root of trust, and signing authority. Redundant power, backup generators, cooling, and fire detection support continuous operation, while retired drives are sanitized and verified or physically destroyed with a shredder and recycled at a secure facility.
The network and perimeter layer controls how traffic moves between users, devices, services, and Google Cloud resources. Google’s global network limits hops across the public internet, which reduces opportunities for interception. Only authorized services and protocols that meet security requirements are allowed to traverse the network, and anything else is automatically dropped.
Segmentation and private connectivity create boundaries so a problem in one area does not spread to others. Organizations can use multiple Virtual Private Cloud (VPC) (opens in a new tab) networks, Shared VPC, VPC Service Controls, Cloud Interconnect, IPsec VPNs, Private Service Connect, and Cloud Service Mesh egress gateways. Traffic flows from customers through Google Front End (GFE) servers to backend systems, where GFE detects and stops malicious requests and distributed denial-of-service (DDoS) attacks. Firewalls and access control lists enforce segregation, logs are examined for exploitation attempts, and Google Cloud Armor can allow, deny, or redirect traffic at the edge.
The data layer protects data at rest, in transit, and in use through encryption and verified code. By default, Google Cloud encrypts customer data at rest in production data centers at the application layer, the storage device layer, or both. Data in transit between devices and GFE is encrypted with strong protocols such as Transport Layer Security (TLS), and virtual machine (VM)-to-VM traffic within a VPC network and between peered VPC networks is encrypted.
Protection for data in use and for keys requires additional controls. Confidential Computing (opens in a new tab) performs computation in cryptographic isolation inside a Trusted Execution Environment (TEE) with hardware isolation and attestation, which helps prevent unauthorized access while data is processed in memory. Binary Authorization for Borg (BAB) helps ensure only authorized code can process data, while customers can use Binary Authorization to enforce their own policies. Key management options include Cloud Key Management Service (Cloud KMS), customer-managed encryption keys (CMEKs), HSM-protected keys, and Cloud External Key Manager (Cloud EKM).
Exam tip: Google Cloud encrypts data at rest and in transit by default, but data in use is not encrypted by default; Confidential Computing protects data in use.
The identity and access layer decides who or what can access resources and under what conditions. Google Cloud follows zero trust, meaning no user or device is inherently trustworthy and every request must be authenticated and authorized. Strong authentication such as multi-factor authentication (MFA) and single sign-on (SSO) is combined with granular policies that consider user role, device posture, and location. The principle of least privilege limits permissions, service accounts receive only necessary permissions, and Workload Identity Federation lets applications access resources securely.
Google also strictly controls its own personnel access to customer data. By default, Google personnel have no access, and to obtain access they must meet conditions including membership in access control lists, regular acknowledgment of data access policies, use of a trusted device, MFA with a Titan Security Key, and tool-based review of justification, role, and context. If required, another qualified Google employee must approve, and customers enrolled in Access Approval must also approve. Access Transparency logs administrative access, and OS Login ties SSH access to Identity and Access Management (IAM) permissions so access ends when roles change.
The operations and monitoring layer finds, prevents, and responds to threats across the environment. Vulnerability management uses automation pipelines that continually reassess patch deployment and flag incorrect or partial deployment. Malware prevention uses web crawling, file detonation, custom static detection, dynamic detection, machine-learning detection, and multiple antivirus engines.
Security monitoring aggregates telemetry in one location for unified analysis. Internal traffic is inspected at many points for suspicious behavior such as botnet connections, and system logs are examined for unusual behavior such as attempts to access customer data. Security Command Center (opens in a new tab), Security Health Analytics, and Mandiant Attack Surface Management help identify misconfigurations and assets across hybrid and multicloud environments. Incident management aligns with NIST SP 800-61, includes staff trained in forensics, tests plans for systems that store customer data, and keeps the security team available 24/7.
How the layers work together describes the ordered chain that protects a request from device to data and back. A user request travels from a device to GFE, where network controls inspect it and TLS protects it in transit. The identity layer then authenticates and authorizes the user or service, the data layer keeps stored data encrypted, and the software supply chain helps ensure only authorized code processes that data. Operations tools collect logs from these layers, detect unusual behavior, and trigger incident response when needed. If one control fails, other layers still limit access, detect the problem, or protect the data, while redundant power and geographically distributed data centers support continuity.
Google’s security posture creates operational and cost advantages because protection is built into the infrastructure and turned on by default. In Google Cloud (opens in a new tab), an organization inherits controls for encryption, network defense, and operations instead of buying and managing many separate tools. This lowers administrative work, reduces risk of outages and breaches, and lets teams spend budget on innovation rather than infrastructure maintenance.
How shared responsibility lowers cost and effort is defined by a clear boundary where Google protects the cloud itself and the customer protects what the customer brings to the cloud. The customer is always responsible for securing its data, while Google secures the underlying infrastructure. This division means a small team does not have to secure data centers, hardware, storage, and networks alone.
Side-by-side comparison of IaaS and SaaS showing Google secures hardware, storage and network in IaaS versus everything except data access in SaaS, while the customer always secures its own data.
| Cloud model | Google secures | Customer secures |
|---|---|---|
| Infrastructure as a service (IaaS) | Hardware, storage, and network | Data, applications, operating systems, and access on top of that foundation |
| Software as a service (SaaS) | Everything except data and its access and usage | Data and how data is accessed and used |
Because Google carries the infrastructure portion, organizations avoid costs for physical security staff, hardware maintenance, and low-level patching. Their work shifts to managing identities, data, and application settings, which directly support business goals. The shared responsibility model (opens in a new tab) therefore reduces both staffing needs and operational risk.
Built-in, always-on controls are security features in Google Cloud that work by default without extra purchase or manual setup. Stored data is encrypted by default at the application layer, the storage device layer, or both, and traffic between devices and Google Front End (GFE) (opens in a new tab) is encrypted with Transport Layer Security (TLS). VM-to-VM traffic inside a Virtual Private Cloud (VPC) network is also encrypted.
These default controls extend across defense-in-depth layers:
Because these controls are part of the platform, fewer add-on tools are needed and automated patching and centralized telemetry reduce misconfigurations that cause downtime.
Exam tip: Google secures the underlying cloud infrastructure, while the customer always secures its own data — built-in encryption does not remove customer responsibility for data access and usage.
Scale advantage in Google Cloud means customers benefit from security investments in people, processes, and technology that most businesses could not afford alone. Google employs a dedicated security team with experts in information security, cryptography, and network security and works with the external research community to find and fix vulnerabilities quickly. The same infrastructure that runs Search, YouTube, and Gmail serves customers, so improvements for Google’s own services directly strengthen customer workloads.
This scale lowers cost because customers do not need large specialist teams or global monitoring operations. Titan hardware security chips (opens in a new tab) and verified software supply chains provide hardware-level trust that is difficult to build independently. Contractual commitments also help customers control how data is processed, and customer data is not used for advertising.
Resilience in Google Cloud comes from redundancy of everything, where no workload depends on a single server, data center, or network connection. Server design, data storage, network connectivity, and software services all use redundant components, and data centers include redundant power, backup generators, cooling, and fire protection. Geographically distributed data centers limit the effect of regional disruptions such as natural disasters or local outages.
The global network (opens in a new tab) improves security and performance by connecting directly to most internet service providers in the world. Data travels the internet in hops and each hop creates a chance for interception, so fewer hops means less exposure. Only authorized services and protocols traverse the network, and traffic flows from customers through GFE to backends where logs are examined. This combination supports high availability and business continuity without customers building backup sites.
Multilayered security, also called defense in depth, is the practice of stacking independent security controls so that if one control fails, another still protects the system. In Google Cloud (opens in a new tab), these layers span physical data centers, custom hardware and software, encryption, the global network, and daily security operations. The business benefit is fewer successful breaches, smaller impact when incidents occur, and stronger continuity and compliance.
How Google’s multilayered defense is organized is a set of physical, technical, and operational controls that share the burden so no single lock or check carries full responsibility. An attacker must defeat several different systems to cause harm, which reduces operational risk because weakness in one area is covered by strength in another. Under the shared responsibility model, Google secures the underlying infrastructure while the customer secures what it brings to the cloud.
The layers include:
Reduced breach likelihood means fewer opportunities for attackers to enter systems or steal data. Layering shrinks the attack surface at every level and blocks unauthorized activity early. Only authorized services and protocols traverse the network while anything else is automatically dropped, and firewalls and access control lists enforce segregation.
Secure design further lowers the chance of compromise. Custom servers omit risky components such as extra video cards and peripheral connectors, vendors are vetted, and development uses source control protections, two-party reviews, and libraries that prevent bugs such as cross-site scripting. Malware defenses use web crawling, file detonation, static and dynamic detection, machine learning, and multiple antivirus engines, while Chrome Enterprise Premium and Gmail Security Sandbox block phishing and suspicious attachments. Because these preventive controls overlap, a single missed vulnerability is far less likely to become a breach.
Reduced impact and business continuity mean services stay available and damage stays contained during failures, attacks, or regional disruptions. Google builds redundancy into server design, data storage, network connectivity, software services, and exception handling. Every critical power component has primary and alternate sources, backup generators can run a data center at full capacity, and cooling and fire suppression protect hardware.
Geographic distribution and intelligent networking also limit impact. Data centers are spread across regions so local outages do not stop global products, and the global network (opens in a new tab) combines owned fiber, public fiber, and undersea cables for low-latency, highly available service. Internal traffic inspection, system log analysis, and correlation systems escalate suspicious activity to staff, and trained responders with forensic tools contain events. This combination resolves incidents faster and resumes normal operations with less downtime and data loss.
Exam tip: Defense in depth reduces both likelihood and impact — prevention layers stop attacks from starting, while redundancy, monitoring, and 24/7 incident response limit harm when something gets through.
Lower organizational risk means less legal, financial, and reputational exposure from security failures and regulatory violations. Google builds independently verified controls and secure defaults into the platform. Customer data at rest is encrypted by default at the application layer, storage device layer, or both, and traffic to GFE is encrypted in transit, so systems can back up data and engineers can support infrastructure without seeing customer content.
These controls help organizations meet needs in finance, government, healthcare, and education. Google Cloud undergoes independent verification against ISO/IEC 27001, ISO/IEC 27017, ISO/IEC 27018, and ISO/IEC 27701, with SOC 2 and SOC 3 reports, plus frameworks such as FedRAMP, BSI C5, and MTCS. Assured Workloads (opens in a new tab) helps enforce regional and regulatory requirements, key management supports rotation and separation of duties, and Google Front End (GFE) (opens in a new tab) filtering and access controls help demonstrate due diligence to auditors.
Google's defense-in-depth model in Google Cloud stacks independent layers across physical data centers, custom hardware, the global network, encryption, identity and access, and security operations so that if one control fails others still protect data and systems. Google secures the underlying cloud infrastructure while customers secure what they put in the cloud, including their data and how it is accessed and used.
Google Cloud encrypts customer data at rest and in transit by default, but data in use is not encrypted by default. Data at rest is encrypted at the application layer, the storage device layer, or both, and data in transit between devices and Google Front End (GFE) is encrypted with strong protocols such as Transport Layer Security (TLS). Confidential Computing protects data in use by performing computation in cryptographic isolation inside a Trusted Execution Environment (TEE) with hardware isolation and attestation.
Under the shared responsibility model Google secures the underlying infrastructure while the customer always secures its own data and how data is accessed and used. Because Google carries the infrastructure portion, organizations avoid costs for physical security staff, hardware maintenance, and low-level patching and shift work to managing identities, data, and application settings. For Infrastructure as a Service (IaaS) Google secures hardware, storage, and network while the customer secures data, applications, operating systems, and access, and for Software as a Service (SaaS) Google secures everything except data and its access and usage.
An enterprise in a regulated industry is preparing to migrate mission-critical workloads and sensitive customer data to Google Cloud. The executive team requires an infrastructure strategy that minimizes organizational risk exposure, prevents single points of failure, and supports strict regulatory compliance requirements.
How does Google Cloud’s defense-in-depth, multilayered security model address these business requirements?
Last updated