AWS Builder Center
Foundational AWS Security Concepts: Shared Responsibility, IAM & AWS Organizations

Foundational AWS Security Concepts: Shared Responsibility, IAM & AWS Organizations

The essential concepts you need for a solid, practical understanding of cloud security.

As someone just exploring the cloud, having knowledge of every service that meets the security needs of a cloud customer in a short time can be quite a task. However, having a solid foundational understanding of the services that support the security of customers' accounts, including their resources and deployed solutions, is a must.
Here, I'll be describing the must-have foundation that Amazon expects you to have to ensure a good experience with its services.
It's safe to say that one of the most important aspects of using cloud services is SECURITY—protecting your data and controlling who can access it. A first step towards achieving this is understanding the Shared Responsibility Model.
Your use of SaaS services, for example, AI chatbots or other products, either as an average user or developer, is not excluded from this conversation. In fact, security and privacy have become major concerns in today's AI-powered world.

Shared Responsibility Model

The Shared Responsibility Model is a cybersecurity principle that divides security responsibilities between a Cloud Service Provider (CSP) and its customer.
It describes how the CSP is responsible for securing the underlying infrastructure that supports its services, known as "Security OF the Cloud," while you remain responsible for how you configure and use the services within the cloud, known as "Security IN the Cloud."
The key principle is that using a cloud service does not eliminate your security responsibilities. If you don't actively configure access restrictions, encryption, and other governance measures and security guardrails, your resources can become insecure regardless of how secure the provider's infrastructure is.
Note that the exact division of responsibilities depends on the cloud service model being used, i.e., IaaS, PaaS, or SaaS.
Generally:
  • IaaS (Infrastructure as a Service): Customers have greater responsibility, including managing operating systems, applications, data, and network configurations.
  • PaaS (Platform as a Service): The provider manages more of the underlying infrastructure and platform, while customers remain responsible for their applications, data, and configurations.
  • SaaS (Software as a Service): The provider manages most of the underlying infrastructure and application, while customers remain responsible for aspects such as data, user access, and service configurations.

What entails "Security OF the Cloud"?

This covers the cloud provider securing the components of the underlying infrastructure that support the vast range of services they offer, ensuring they remain available and operate securely around the clock.
It covers:
  • Physical security: Data centers with controlled, need-based access, 24/7 security personnel, access logging and reviews, video surveillance, and secure media handling, including disk sanitization and destruction.
  • Hardware infrastructure: Servers, storage devices, and other physical appliances that AWS relies on to deliver its services.
  • Software infrastructure: The underlying software, including virtualization software and infrastructure components that support AWS services.
  • Network infrastructure: Routers, switches, load balancers, firewalls, and cabling. AWS also monitors its network boundaries, secures access points, and maintains redundant infrastructure.

What entails "Security IN the Cloud"?

This is the responsibility division where customers are responsible for what they use or store in the cloud. It covers protecting data, applications, identities, and resources deployed in their cloud environments.
Customers are generally responsible for:
  • Data security: Protecting sensitive information through encryption, access controls, and appropriate data classification.
  • Identity and Access Management (IAM): Managing user accounts, permissions, authentication, and multifactor authentication (MFA).
  • Application security: Developing, configuring, and maintaining secure applications.
  • Configuration management: Ensuring cloud resources are configured securely and avoiding publicly exposed resources.
  • Operating system security: Patching and securing guest operating systems running on instances.
  • Network security configuration: Configuring firewalls, security groups, and network access rules within their cloud environment.
  • Compliance: Meeting applicable regulatory and organizational requirements for the data and workloads they manage.

Security Within Your AWS Account

Your AWS account is your go-to portal for accessing the vast range of services available in AWS. Hence, it helps to know and adopt best practices for properly managing your account to ensure a good user experience and maintain security.

Identity and Access Management — AWS IAM

Remember how you can have different user accounts on a single PC and also decide which of them—users or groups—gets access to important files on your storage?
Well, that's Identity and Access Management at work, and the same concept can be applied to your AWS account through the AWS Identity and Access Management (IAM) service.
IAM is a service that helps answer two fundamental security questions:
  • Authentication: Who are you?
  • Authorization: What are you allowed to do?
IAM is an AWS service offered at no additional charge.

Fine-Grained Access vs. Granular Control

Choosing who gets access and determining the extent of the actions they can perform on a resource focuses on two important IAM concepts: Fine-Grained Access and Granular Control.

Fine-Grained Access

Fine-grained access means giving users only the specific permissions they need to perform their tasks, rather than giving them access to everything.
Example:
Imagine you work in a company that uses cloud storage.
  • John works in the Finance department. He can access financial documents.
  • Mary works in HR. She can access employee records.
  • Peter works in IT. He can manage cloud infrastructure.
Each person gets access only to the resources they need.
In simple terms: Not everyone gets access to everything. Access is carefully assigned based on individual roles and responsibilities.

Granular Control

Granular control means having the ability to manage permissions in detail, deciding exactly what a user can or cannot do within a particular resource.
Example:
Imagine John has access to a financial document.
You can decide that:
  • John can view the document.
  • John can edit the document.
  • John cannot delete the document.
  • John cannot share the document with others.
Although John has access to the document, his actions are restricted to what he is permitted to do.
In simple terms: You don't just decide who gets access; you also decide exactly what they can do with that access.
Bearing these concepts in mind enables an administrator to confidently implement the security principle of least privilege.
The principle of least privilege promotes granting users only the minimum permissions necessary to perform their assigned tasks.

Essential Components of AWS IAM

IAM centrally manages access to AWS resources and services. It provides granular control over access, including the ability to specify exactly which API calls a user or entity is authorized to make to each service.
Whether you use the AWS Management Console, AWS Command Line Interface (CLI), or AWS Software Development Kits (SDKs), interactions with AWS services involve API calls.
The components of AWS IAM that help secure a cloud customer's account and ensure the proper usage of deployed solutions and provisioned resources can be categorized into two mechanisms: Identity Mechanisms and Access Management Mechanisms.
The components of AWS IAM occupying the identity mechanism category are IAM Users, IAM Groups, and IAM Roles, while IAM Policies form the foundation of the access management mechanism.



IAM User

An IAM user is an identity created within an AWS account to represent a person or application that needs to interact with AWS services.
An IAM user can have credentials that allow it to authenticate and make API calls to AWS services within the account.
Best practices to follow when creating users:
  • Each IAM user must have a unique name within the AWS account.
  • Users should have appropriate security credentials for authentication, and those credentials must be stored securely and never shared.
  • Avoid creating long-term IAM user credentials when temporary credentials through IAM roles or identity federation can be used instead.
For human users, AWS recommends using centralized identity management solutions such as IAM Identity Center where appropriate.

IAM Group

An IAM group is a collection of IAM users with the same permissions and access requirements.
It is created to simplify permission management by allowing administrators to assign permissions to multiple users at once rather than assigning permissions individually, which can become inefficient and difficult to manage, especially in organizations with many users.
In other words, its purpose is to simplify permission management for multiple users.
Note: IAM groups do not have security credentials of their own, and IAM groups cannot be nested within other groups.

IAM Role

One major difference between an IAM role and an IAM user is that an IAM role provides temporary security credentials when assumed, whereas IAM users can have long-term security credentials.
An IAM role is an identity that provides temporary permissions to users, applications, or AWS services that need to access specific AWS resources.
For example, an EC2 instance can assume an IAM role to access a particular S3 bucket without requiring hardcoded access keys.
When creating an IAM role in AWS, you typically configure two important things:
  • Trust Policy (Who can assume the role): Defines which AWS service, user, or account is trusted to assume the role.
  • Permissions Policy (What the role can do): Defines which AWS resources and actions the role can access.
IAM roles can also facilitate cross-account access, allowing trusted identities from one AWS account to access resources in another account when properly configured.

IAM Policy

An IAM policy is a formal statement of permissions that determines what actions an entity (identity or resource) is allowed or denied to perform.
Policies can be attached to IAM users, groups, and roles, while resource-based policies are attached directly to supported AWS resources.
There are two commonly encountered policy types:
  • Identity-based policies: Attached to IAM identities (users, groups, or roles) to define what actions they can perform on which resources.
  • Resource-based policies: Attached to resources to define who can access them and what actions they can perform.
Both are typically written as JSON (JavaScript Object Notation) documents.



Authentication vs. Authorization

These are two validation and access-control processes that answer different questions.

Authentication — Who are you?

Authentication is the process of verifying the identity of a user or entity.
It answers the question, "Who are you?" by requesting credentials, such as a username and password, access keys, or temporary security credentials, which provide evidence that the claimed identity is genuine.

Authorization — What are you allowed to access or do?

Authorization is the process of determining the access rights and permissions of an authenticated user or entity.
It answers the question, "I know who you are, but what are you allowed to do?" by granting or restricting access to specific resources and actions based on the applicable permissions.
To grant permissions to a user, group, or role, you must create an IAM policy or use an existing policy in the account.
IAM does not automatically grant users permissions to AWS resources simply because their identities have been created.

Access Control Principles



There are two important access-control principles to understand:
  • Implicit Deny: If an action is not explicitly allowed by an applicable policy, access is denied by default.
  • Explicit Deny: If an applicable policy explicitly denies an action, that denial overrides any Allow permission.
For example, if a user has permission to access an S3 bucket but another applicable policy explicitly denies that access, the explicit Deny takes precedence.
When creating IAM policies, it is best practice to follow the security advice of granting least privilege. Determine what users need to be able to do, then craft policies that allow them to perform only those tasks.
Start with a minimum set of permissions and grant additional permissions as necessary.
Note: IAM is a global AWS service. Its configuration is not region-specific, and IAM identities and policies apply across AWS Regions.

AWS Organizations

So far, the discussion has focused on security within an individual AWS account. Here, we'll briefly describe a service that helps in securely managing multiple AWS accounts: AWS Organizations.
AWS Organizations is a management and governance service that enables you to consolidate multiple AWS accounts into an organization that can be centrally managed.
A useful feature of the service is consolidated billing, which allows the organization to combine billing across its accounts.
Imagine a company with four AWS accounts serving different purposes:
  • Management Account: Handles organization management and consolidated billing.
  • Development Account: Used for building and testing applications.
  • Production Account: Runs live applications.
  • Security Account: Handles security monitoring and auditing.
AWS Organizations allows the company to group and govern these accounts centrally.
The organization is managed through its management account.
Each member account remains a distinct AWS account, with its own resources and IAM configurations.
It is important to note that AWS Organizations does not automatically give the management account unrestricted access to every resource in every member account.
Central administration and direct access to resources are different things. Organizations govern accounts; they do not replace IAM within those accounts.

Key Security Features AWS Organizations Provides

  • Grouping accounts into Organizational Units (OUs): This lets you group similar accounts together, much like folders, and apply policies to the entire group. For instance, you can place accounts that require strict compliance into one OU and apply restrictions to help enforce organizational security standards.
  • Integration with IAM: AWS Organizations works alongside IAM to establish permission boundaries across multiple accounts. A user or role must have the necessary IAM permissions, and applicable organization-level restrictions must also allow the requested action.
  • Service Control Policies (SCPs): SCPs establish the maximum permissions available to identities in member accounts. They act as guardrails that restrict the actions users and roles can perform, including account administrators. If an SCP blocks an action, a local administrator cannot override that restriction simply by granting additional IAM permissions.
Important: SCPs do not grant permissions; they only restrict the maximum available permissions. Also, SCPs do not apply to principals in the organization's management account or service-linked roles.
In essence, AWS Organizations helps businesses maintain centralized governance and enforce security standards across multiple AWS accounts without necessarily managing each account independently.
And that's the foundation!
Understanding the Shared Responsibility Model, IAM, and AWS Organizations gives you a solid starting point for navigating AWS security.
As you progress, you can build on this foundation by exploring other security services such as AWS CloudTrail, AWS Config, Amazon GuardDuty, AWS Security Hub, AWS Key Management Service (KMS), and AWS Secrets Manager.
Remember, cloud security is not just about knowing which services exist; it's about understanding how to use them together to protect your resources, data, and deployed solutions.
Any opinions in this article are those of the individual author and may not reflect the opinions of AWS.
Enjoyed reading this content? Let the author know!

Your likes, comments, shares, and saves help creators reach more builders.

Loading recommendations

Loading article