Security best practices for Amazon S3

S3 Object Ownership is an Amazon S3 bucket-level setting that you can use to control ownership of objects uploaded to your bucket and to disable or enable ACLs. By default, Object Ownership is set to the Bucket owner enforced setting and all ACLs are disabled. When ACLs are disabled, the bucket owner owns all the objects in the bucket and manages access to data exclusively using access management policies.

A majority of modern use cases in Amazon S3 no longer require the use of access control lists (ACLs). We recommend that you disable ACLs, except in unusual circumstances where you must control access for each object individually. To disable ACLs and take ownership of every object in your bucket, apply the bucket owner enforced setting for S3 Object Ownership. When you disable ACLs, you can easily maintain a bucket with objects uploaded by different AWS accounts.

When ACLs are disabled access control for your data is based on policies, such as the following:

Disabling ACLs simplifies permissions management and auditing. ACLs are disabled for new buckets by default. You can also disable ACLs for existing buckets. If you have an existing bucket that already has objects in it, after you disable ACLs, the object and bucket ACLs are no longer part of the access-evaluation process. Instead, access is granted or denied on the basis of policies.

Before you disable ACLs, make sure that you do the following:

After you disable ACLs, the following behaviors occur:

For more information, see Controlling ownership of objects and disabling ACLs for your bucket.

Unless you explicitly require anyone on the internet to be able to read or write to your S3 bucket, make sure that your S3 bucket is not public. The following are some of the steps that you can take to block public access:

For more information, see Identity and Access Management for Amazon S3.

When granting permissions, you decide who is getting what permissions to which Amazon S3 resources. You enable specific actions that you want to allow on those resources. Therefore, we recommend that you grant only the permissions that are required to perform a task. Implementing least privilege access is fundamental in reducing security risk and the impact that could result from errors or malicious intent.

The following tools are available to implement least privilege access:

For guidance on what to consider when choosing one or more of the preceding mechanisms, see Identity and Access Management for Amazon S3.

In order for applications running on Amazon EC2 or other AWS services to access Amazon S3 resources, they must include valid AWS credentials in their AWS API requests. We recommend not storing AWS credentials directly in the application or Amazon EC2 instance. These are long-term credentials that are not automatically rotated and could have a significant business impact if they are compromised.

Instead, use an IAM role to manage temporary credentials for applications or services that need to access Amazon S3. When you use a role, you don't have to distribute long-term credentials (such as a username and password or access keys) to an Amazon EC2 instance or AWS service, such as AWS Lambda. The role supplies temporary permissions that applications can use when they make calls to other AWS resources.

For more information, see the following topics in the IAM User Guide:

You have the following options for protecting data at rest in Amazon S3:

You can use HTTPS (TLS) to help prevent potential attackers from eavesdropping on or manipulating network traffic by using person-in-the-middle or similar attacks. We recommend allowing only encrypted connections over HTTPS (TLS) by using the aws:SecureTransport condition in your Amazon S3 bucket policies.

Also consider implementing ongoing detective controls by using the s3-bucket-ssl-requests-only managed AWS Config rule.

With S3 Object Lock, you can store objects by using a "Write Once Read Many" (WORM) model. S3 Object Lock can help prevent accidental or inappropriate deletion of data. For example, you can use S3 Object Lock to help protect your AWS CloudTrail logs.

For more information, see Locking objects with Object Lock.

S3 Versioning is a means of keeping multiple variants of an object in the same bucket. You can use versioning to preserve, retrieve, and restore every version of every object stored in your bucket. With versioning, you can easily recover from both unintended user actions and application failures.

Also consider implementing ongoing detective controls by using the s3-bucket-versioning-enabled managed AWS Config rule.

For more information, see Retaining multiple versions of objects with S3 Versioning.

Although Amazon S3 stores your data across multiple geographically diverse Availability Zones by default, compliance requirements might dictate that you store data at even greater distances. With S3 Cross-Region Replication (CRR), you can replicate data between distant AWS Regions to help satisfy these requirements. CRR enables automatic, asynchronous copying of objects across buckets in different AWS Regions. For more information, see Replicating objects within and across Regions.

Also consider implementing ongoing detective controls by using the s3-bucket-replication-enabled managed AWS Config rule.

A virtual private cloud (VPC) endpoint for Amazon S3 is a logical entity within a VPC that allows connectivity only to Amazon S3. VPC endpoints can help prevent traffic from traversing the open internet.

VPC endpoints for Amazon S3 provide multiple ways to control access to your Amazon S3 data:

For more information, see Controlling access from VPC endpoints with bucket policies.

Several managed AWS security services can help you identify, assess, and monitor security and compliance risks for your Amazon S3 data. These services can also help you protect your data from those risks. These services include automated detection, monitoring, and protection capabilities that are designed to scale from Amazon S3 resources for a single AWS account to resources for organizations spanning thousands of accounts.

For more information, see Monitoring data security with managed AWS security services.

Identification of your IT assets is a crucial aspect of governance and security. You need to have visibility of all your Amazon S3 resources to assess their security posture and take action on potential areas of weakness. To audit your resources, we recommend doing the following:

Monitoring is an important part of maintaining the reliability, security, availability, and performance of Amazon S3 and your AWS solutions. AWS provides several tools and services to help you monitor Amazon S3 and your other AWS services. For example, you can monitor Amazon CloudWatch metrics for Amazon S3, particularly the PutRequests, GetRequests, 4xxErrors, and DeleteRequests metrics. For more information, see Monitoring metrics with Amazon CloudWatch and Logging and monitoring in Amazon S3.

For a second example, see Example: Amazon S3 Bucket Activity. This example describes how to create a CloudWatch alarm that is triggered when an Amazon S3 API call is made to PUT or DELETE a bucket policy, a bucket lifecycle, or a bucket replication configuration, or to PUT a bucket ACL.

Server access logging provides detailed records of the requests that are made to a bucket. Server access logs can assist you in security and access audits, help you learn about your customer base, and understand your Amazon S3 bill. For instructions on enabling server access logging, see Logging requests with server access logging.

Also consider implementing ongoing detective controls by using the s3-bucket-logging-enabled AWS Config managed rule.

AWS CloudTrail provides a record of actions taken by a user, a role, or an AWS service in Amazon S3. You can use information collected by CloudTrail to determine the following:

For example, you can identify CloudTrail entries for PUT actions that affect data access, in particular PutBucketAcl, PutObjectAcl, PutBucketPolicy, and PutBucketWebsite.

When you set up your AWS account, CloudTrail is enabled by default. You can view recent events in the CloudTrail console. To create an ongoing record of activity and events for your Amazon S3 buckets, you can create a trail in the CloudTrail console. For more information, see Logging data events in the AWS CloudTrail User Guide.

When you create a trail, you can configure CloudTrail to log data events. Data events are records of resource operations performed on or within a resource. In Amazon S3, data events record object-level API activity for individual buckets. CloudTrail supports a subset of Amazon S3 object-level API operations, such as GetObject, DeleteObject, and PutObject. For more information about how CloudTrail works with Amazon S3, see Logging Amazon S3 API calls using AWS CloudTrail. In the Amazon S3 console, you can also configure your S3 buckets to Enabling CloudTrail event logging for S3 buckets and objects.

AWS Config provides a managed rule (cloudtrail-s3-dataevents-enabled) that you can use to confirm that at least one CloudTrail trail is logging data events for your S3 buckets. For more information, see cloudtrail-s3-dataevents-enabled in the AWS Config Developer Guide.

Several of the best practices listed in this topic suggest creating AWS Config rules. AWS Config helps you to assess, audit, and evaluate the configurations of your AWS resources. AWS Config monitors resource configurations so that you can evaluate the recorded configurations against the desired secure configurations. With AWS Config, you can do the following:

Using AWS Config can help you simplify compliance auditing, security analysis, change management, and operational troubleshooting. For more information, see Setting Up AWS Config with the Console in the AWS Config Developer Guide. When specifying the resource types to record, ensure that you include Amazon S3 resources.

For an example of how to use AWS Config, see How to Use AWS Config to Monitor for and Respond to Amazon S3 Buckets Allowing Public Access on the AWS Security Blog.

S3 Storage Lens is a cloud-storage analytics feature that you can use to gain organization-wide visibility into object-storage usage and activity. S3 Storage Lens also analyzes metrics to deliver contextual recommendations that you can use to optimize storage costs and apply best practices for protecting your data.

With S3 Storage Lens, you can use metrics to generate summary insights, such as finding out how much storage you have across your entire organization or which are the fastest-growing buckets and prefixes. You can also use S3 Storage Lens metrics to identify cost-optimization opportunities, implement data-protection and access-management best practices, and improve the performance of application workloads.

For example, you can identify buckets that don't have S3 Lifecycle rules to abort incomplete multipart uploads that are more than 7 days old. You can also identify buckets that aren't following data-protection best practices, such as using S3 Replication or S3 Versioning. For more information, see Understanding Amazon S3 Storage Lens.

We recommend that you regularly check the security advisories posted in Trusted Advisor for your AWS account. In particular, look for warnings about Amazon S3 buckets with "open access permissions." You can do this programmatically by using describe-trusted-advisor-checks.

Further, actively monitor the primary email address that's registered to each of your AWS accounts. AWS uses this email address to contact you about emerging security issues that might affect you.

AWS operational issues with broad impact are posted on the AWS Health Dashboard - Service health. Operational issues are also posted to individual accounts through the AWS Health Dashboard. For more information, see the AWS Health documentation.

Amazon GuardDuty is a threat-detection service that continuously monitors your AWS accounts and workloads for malicious activity and delivers detailed security findings for visibility and remediation.

With the S3 protection feature in GuardDuty, you can configure GuardDuty to analyze AWS CloudTrail management and data events for your Amazon S3 resources. GuardDuty then monitors those events for malicious and suspicious activity. To inform the analysis and identify potential security risks, GuardDuty uses threat-intelligence feeds and machine learning.

GuardDuty can monitor different kinds of activity for your Amazon S3 resources. For example, CloudTrail management events for Amazon S3 include bucket-level operations, such as ListBuckets, DeleteBucket, and PutBucketReplication. CloudTrail data events for Amazon S3 include object-level operations, such as GetObject, ListObjects, and PutObject. If GuardDuty detects anomalous or potentially malicious activity, it generates a finding to notify you.

For more information, see Amazon S3 Protection in Amazon GuardDuty in the Amazon GuardDuty User Guide.

Amazon Detective simplifies the investigative process and helps you conduct faster, more effective security investigations. Detective provides prebuilt data aggregations, summaries, and context that can help you analyze and assess the nature and extent of possible security issues.

Detective automatically extracts time-based events, such as API calls from AWS CloudTrail and Amazon VPC Flow Logs for your AWS resources. It also ingests findings generated by Amazon GuardDuty. Detective then uses machine learning, statistical analysis, and graph theory to generate visualizations that help you conduct effective security investigations more quickly.

These visualizations provide a unified, interactive view of resource behaviors and the interactions between them over time. You can explore this behavior graph to examine potentially malicious actions, such as failed login attempts or suspicious API calls. You can also see how these actions affect resources, such as S3 buckets and objects.

For more information, see the Amazon Detective Administration Guide.

AWS Identity and Access Management Access Analyzer (IAM Access Analyzer) can help you identify resources that are shared with an external entity. You can also use IAM Access Analyzer to validate IAM policies against policy grammar and best practices, and generate IAM policies based on access activity in your AWS CloudTrail logs.

IAM Access Analyzer uses logic-based reasoning to analyze resource policies in your AWS environment, such as bucket policies. With IAM Access Analyzer for S3, you're alerted when an S3 bucket is configured to allow access to anyone on the internet or other AWS accounts, including accounts outside your organization. For example, IAM Access Analyzer for S3 can report that a bucket has read or write access provided through a bucket access control list (ACL), a bucket policy, a Multi-Region Access Point policy, or an access point policy. For each public or shared bucket, you receive findings that indicate the source and level of public or shared access. With these findings, you can take immediate and precise corrective action to restore bucket access to what you intended.

For more information, see Reviewing bucket access using IAM Access Analyzer for S3.

Amazon Macie is a security service that discovers sensitive data by using machine learning and pattern matching. Macie provides visibility into data security risks, and enables automated protection against those risks. With Macie, you can automate the discovery and reporting of sensitive data in your Amazon S3 data estate to gain a better understanding of the data that your organization stores in S3.

To detect sensitive data with Macie, you can use built-in criteria and techniques that are designed to detect a large and growing list of sensitive data types for many countries and regions. These sensitive data types include multiple types of personally identifiable information (PII), financial data, and credentials data. You can also use custom criteria that you define—regular expressions that define text patterns to match and, optionally, character sequences and proximity rules that refine the results.

If Macie detects sensitive data in an S3 object, Macie generates a security finding to notify you. This finding provides information about the affected object, the types and number of occurrences of the sensitive data that Macie found, and additional details to help you investigate the affected S3 bucket and object. For more information, see the Amazon Macie User Guide.

AWS Security Hub is a security-posture management service that performs security best-practice checks, aggregates alerts and findings from multiple sources into a single format, and enables automated remediation.

Security Hub collects and provides security findings data from integrated AWS Partner Network security solutions and AWS services, including Amazon Detective, Amazon GuardDuty, IAM Access Analyzer, and Amazon Macie. It also generates its own findings by running continuous, automated security checks based on AWS best practices and supported industry standards.

Security Hub then correlates and consolidates findings across providers to help you prioritize and process the most significant findings. It also provides support for custom actions, which you can use to invoke responses or remediation actions for specific classes of findings.

With Security Hub, you can assess the security and compliance status of your Amazon S3 resources, and you can do so as part of a broader analysis of your organization's security posture in individual AWS Regions and across multiple Regions. This includes analyzing security trends and identifying the highest-priority security issues. You can also aggregate findings from multiple AWS Regions, and monitor and process aggregated findings data from a single Region.

For more information, see Amazon Simple Storage Service controls in the AWS Security Hub User Guide.