As a Cloud Security Architect, DevSecOps Lead, and Terraform Specialist, I'm pleased to present today's comprehensive summary of new AWS features and service updates, focusing on their security implications and Infrastructure as Code (IaC) availability.

Category: Analytics

Amazon Redshift has expanded its long-term system table retention with Amazon S3 Tables to additional AWS Regions, including Asia Pacific (New Zealand), Canada West (Calgary), AWS GovCloud (US-East), and AWS GovCloud (US-West). This feature allows customers to retain system table data beyond the standard 7-day limit, with AWS automatically writing this data to S3 Tables in Apache Iceberg format and managing partitioning, compaction, and retention. This is a significant enhancement for auditability and compliance, enabling organizations to meet stringent data retention policies and perform historical analysis of cluster operations. The ability to query this data via Redshift, Athena, DuckDB, and other Iceberg-compatible engines provides flexibility for security and operational insights. Architects should leverage this for robust compliance reporting and forensic analysis capabilities. *Terraform Support*: Pending provider release (use AWS CLI/SDK or custom provisioner in the interim).

Category: Compute

AWS Lambda has extended its recursive loop detection mechanism to include Amazon EventBridge Custom event buses. This critical security feature automatically detects and stops unintended recursive invocations of Lambda functions triggered by misconfigurations or code defects, preventing runaway costs and potential denial-of-service scenarios. Enabled by default at the account level, it requires no explicit configuration for existing or new functions. Upon detection, an AWS Health Dashboard notification is sent with troubleshooting guidance. For intentional recursive patterns, the PutFunctionRecursionConfig API allows disabling detection for specific functions. This enhancement significantly improves the resilience and security posture of event-driven architectures. *Terraform Support*: Supported in Terraform AWS Provider: aws_lambda_function (argument recursion_config) in v5.27.0+.

Category: Database

PostgreSQL 19 RC1 is now available in the Amazon RDS Database Preview Environment. This pre-release version focuses on stability, incorporating bug fixes and improvements from Beta 4, such as enhanced error handling for complex text-search expressions, hardened parallel autovacuum, and improved reliability of GIN index maintenance. This allows organizations to proactively evaluate the new PostgreSQL version, test application compatibility, and assess any security implications or performance improvements before its general availability. Utilizing the preview environment is a best practice for maintaining production stability and ensuring a smooth upgrade path. *Terraform Support*: Supported in Terraform AWS Provider: aws_db_instance (argument engine_version) in v5.55.0+.

Oracle Database@AWS now supports advertising aggregate CIDR ranges and a default route (0.0.0.0/0) over ODB peering connections. This significantly reduces operational overhead by eliminating the need to explicitly advertise each network's CIDR for connectivity to the ODB network. A single aggregate or default route can now cover multiple networks reachable through the peered VPC, including those connected via AWS Transit Gateway. From a security perspective, while simplifying network management, architects must carefully manage the scope of default routes to prevent unintended network access and ensure proper segmentation, especially when extending connectivity to on-premises environments. *Terraform Support*: Pending provider release (use AWS CLI/SDK or custom provisioner in the interim).

Category: Data Orchestration

Amazon MWAA now integrates with the Agent Toolkit for AWS, introducing built-in AI agent skills for authoring, debugging, and testing MWAA workflows, as well as upgrading MWAA environments. These four skills enable AI coding assistants to interact directly with MWAA using natural language, automating the creation and maintenance of data pipelines. For example, describing an ETL pipeline in plain language can generate a working DAG with dependencies and error handling. This accelerates development and reduces manual effort. From a DevSecOps perspective, while AI can streamline secure pipeline design, it's crucial to implement robust security controls around AI agent permissions and to rigorously review and validate AI-generated code for potential vulnerabilities or misconfigurations. *Terraform Support*: Pending provider release (use AWS CLI/SDK or custom provisioner in the interim).

Category: Machine Learning

Amazon Bedrock's Web Search capability now offers enhanced control with support for domain filtering (up to 100 allowed_domains or blocked_domains via the filters parameter) and optional user location parameters for geography-specific results. These controls are configured directly on the web_search tool in API requests and are available in key AWS regions. This feature is critical for data governance and compliance, allowing organizations to restrict AI agents from accessing or generating content from specific sources, or to tailor responses based on user location. Security architects can leverage these filters to prevent exposure to undesirable content, enforce data sovereignty, and ensure AI agent responses adhere to organizational policies. *Terraform Support*: Pending provider release (use AWS CLI/SDK or custom provisioner in the interim).

Category: Messaging

AWS End User Messaging has introduced a guided wizard to orchestrate the entire video capture process for RCS agent registration. This wizard simplifies the submission of compliant videos by guiding users through each step, sending appropriate messages to a test device, and ensuring the video meets regulatory requirements the first time. This feature is available on all RCS registration forms requiring video submission. From a security and compliance standpoint, this streamlines a critical identity verification step, reducing the likelihood of non-compliant submissions and accelerating the secure onboarding of RCS agents. *Terraform Support*: Pending provider release (use AWS CLI/SDK or custom provisioner in the interim).

Category: Migration

AWS Transform for migrations is now available in the Asia Pacific (Singapore) Region. This agentic AI-powered service accelerates cloud migration, application modernization, and continuous tech debt reduction. Its regional expansion provides customers with data residency within Singapore and lower latency to local source infrastructure. AWS Transform automates the entire migration lifecycle, from discovery and wave planning to network conversion and launching test/cutover instances, unifying teams, tools, and workflows. For security, it's paramount to ensure that the AI agents and the service itself operate with least privilege, and that all data transferred and transformed during migration adheres to security and compliance policies. *Terraform Support*: Pending provider release (use AWS CLI/SDK or custom provisioner in the interim).

Category: Monitoring

CloudWatch alarms can now include alarm tags in the messages sent to Amazon SNS, AWS Lambda, and Amazon EventBridge when an alarm changes state. This feature is invaluable for customers routing numerous alarms through shared notification pipelines, enabling direct identification of the owning team, environment, or service from the message payload. The setting is off by default and configurable per alarm, allowing inclusion of all tags, selected tag keys, or all tags except selected keys. This is supported for metric, composite, and log alarms. From a DevSecOps perspective, this significantly enhances automation and incident response by providing immediate context, allowing for more intelligent routing and processing of security alerts based on tag metadata. *Terraform Support*: Supported in Terraform AWS Provider: aws_cloudwatch_metric_alarm (arguments tags_on_actions_enabled, tags_on_actions_include_keys, tags_on_actions_exclude_keys) in v5.60.0+.

Category: Management & Governance

CloudWatch pipelines now offer conditional log routing and chained parsers. Conditional log routing allows users to define rules to send log events to different CloudWatch log groups based on conditions evaluated against processed fields, enabling tailored retention and access controls. Chained parsers facilitate the sequential application of multiple parser processors, where subsequent parsers can consume fields generated by earlier ones (e.g., parsing a JSON envelope then a Grok pattern on an extracted field). These capabilities streamline the structuring of diverse log formats and their routing. For security, this enables granular segregation of security-relevant logs, facilitating targeted analysis, faster threat detection, and adherence to data residency and access control policies. *Terraform Support*: Pending provider release (use AWS CLI/SDK or custom provisioner in the interim).

Category: Observability

The Amazon Managed Service for Prometheus collector, a fully-managed agentless collector for Prometheus metrics, is now available in Asia Pacific (New Zealand). This expansion allows customers in New Zealand to run Prometheus-compatible monitoring workloads in-region, benefiting from lower latency for local workloads and meeting data residency requirements. The agentless collector simplifies operational overhead by eliminating the need to run and operate collection agents, reducing the attack surface and enhancing the security posture of the monitoring infrastructure. Collected metrics are ingested into an AMP workspace for querying and analysis. *Terraform Support*: Supported in Terraform AWS Provider: aws_prometheus_scraper in v5.10.0+.

Category: Security, Identity, & Compliance

AWS Security Hub CSPM now supports the Cyber Risk Institute (CRI) profile version 2.2, introducing 404 new automated security controls. This integration is specifically tailored for the financial services sector, harmonizing cybersecurity and regulatory expectations. It enables organizations to perform a single assessment and report against multiple regulations, significantly reducing audit fatigue and streamlining compliance efforts. The new standard is available across all AWS Regions where Security Hub CSPM operates, including AWS GovCloud (US) and China Regions, and can be efficiently enabled using Security Hub's central configuration capabilities. This is a critical update for financial institutions aiming to automate and strengthen their compliance posture. *Terraform Support*: Supported in Terraform AWS Provider: aws_securityhub_standards_subscription (argument standards_arn) in v2.37.0+.

Category: Storage

Amazon S3 Metadata and its 'Annotations' feature are now generally available in six additional AWS Regions. S3 Metadata provides automated, near real-time metadata updates for S3 objects, offering a comprehensive and queryable view of data. The Annotations feature extends this by allowing users to attach rich business context (up to 1 GB per object) in JSON, XML, or YAML formats, which can then be surfaced and queried within S3 Metadata annotation tables. This significantly enhances data governance, discovery, and classification capabilities. From a security perspective, leveraging rich metadata and annotations enables more granular access policies, easier identification of sensitive data, and improved data lifecycle management, crucial for compliance with data protection regulations. *Terraform Support*: Pending provider release (use AWS CLI/SDK or custom provisioner in the interim).

Category: Streaming

Amazon MSK Connect, which provides fully managed Kafka Connect clusters, and AWS Glue Schema Registry, a serverless and free feature for validating and controlling streaming data schemas, are now available in additional AWS regions. MSK Connect simplifies the deployment, monitoring, and scaling of connectors that move data between Apache Kafka and external systems. The AWS Glue Schema Registry acts as a centralized repository for data format and structure, supporting Apache Avro, JSON, and Protobuf schemas, and integrates with various AWS streaming services. This combination reduces operational overhead, ensures data quality, and enhances compliance by enforcing data contracts and standardizing data formats across decoupled applications. *Terraform Support*: Supported in Terraform AWS Provider: aws_mskconnect_connector in v4.22.0+.

Announcements in this summary