As a Cloud Security Architect and DevSecOps Lead, I've reviewed today's AWS updates, focusing on their security implications, architectural considerations, and Terraform availability. These features offer significant advancements in operational efficiency, cost management, and security posture across various domains.

Compute & Migration Highlights

The latest update to AWS Transform for .NET introduces the capability to automatically generate unit tests for modernized code. This feature, opt-in via the AWS Toolkit for Visual Studio, assesses applications for testability, identifies key classes, and generates corresponding unit test code. This significantly reduces manual effort in .NET Framework to modern .NET migrations, ensuring pre-existing test coverage. From a security and architectural standpoint, integrating automated testing early in the migration process helps maintain code quality, catch regressions, and identify potential vulnerabilities before deployment, contributing to a more secure and stable application. Compliance benefits include demonstrating robust testing practices. *Terraform Support*: Pending provider release (use AWS CLI/SDK or custom provisioner in the interim).

AWS Lambda durable functions now integrate with Pydantic AI to automatically checkpoint the progress of AI agents. Each model and tool call is treated as a durable execution step, allowing agents to resume from the last completed step if a Lambda execution is interrupted. This is crucial for long-running or expensive AI workflows, preventing repeated token costs and avoiding unwanted side-effects like duplicate billing. Security-wise, this reduces the attack surface by ensuring predictable execution and preventing resource exhaustion from repeated restarts. Architecturally, it enables more resilient and cost-efficient AI agent orchestration. *Terraform Support*: Supported for the underlying aws_lambda_function resource in Terraform AWS Provider: aws_lambda_function (general Lambda function configuration) in v0.1.0+ up to v5.31.0. The Pydantic AI integration is primarily an SDK/runtime feature.

AWS Lambda's recursive loop detection is now supported for functions running in Europe Sovereign Cloud. This feature automatically detects and stops recursive invocations between Lambda functions and other supported services, preventing unexpected billing and potential denial-of-service scenarios caused by unintended loops. It's enabled by default for functions using a supported SDK version and sends notifications via the AWS Health Dashboard. This is a critical security and operational guardrail, especially in regulated environments, ensuring cost predictability and service stability. *Terraform Support*: Supported in Terraform AWS Provider: aws_lambda_function (argument recursion_config) in v5.47.0+ up to v5.70.0.

AWS Outposts now offers second-generation single-rack units, a self-contained 42U rack integrating compute, storage, and networking. Designed for workloads requiring low latency, local data processing, and data residency in space and power-constrained locations, it supports up to 2,688 vCPU and 100 TB of Amazon EBS storage with the latest EC2 instances. This extends the consistent AWS experience, including security controls, APIs, and governance, to on-premises environments. Security benefits include a unified security posture across hybrid deployments and meeting strict data residency requirements. Architecturally, it enables true hybrid cloud strategies for sensitive or performance-critical applications. *Terraform Support*: Supported in Terraform AWS Provider: aws_outposts_outpost (for configuring Outpost instance types and physical properties) in v3.1.0+ up to v5.x.x (or later).

Amazon ECS now extends the ecs:task-cpu and ecs:task-memory IAM condition keys to the RunTask and StartTask APIs. This provides a unified mechanism for controlling resource allocation (CPU and memory) for ECS tasks, regardless of whether they are launched via service creation/update or direct task execution. From a security perspective, this enables fine-grained access control, preventing unauthorized resource consumption or privilege escalation. Architecturally, it ensures consistent governance and cost control across all task launch methods, enhancing operational stability and predictability. *Terraform Support*: Supported in Terraform AWS Provider: aws_iam_policy (for condition blocks within IAM policies) in v2.0.0+ up to v5.30.0.

Application Integration & Messaging

Amazon API Gateway now supports configurable delivery destinations and larger log events (up to 1 MB) for REST API execution logs. You can route these logs to your own Amazon CloudWatch Logs log groups, Amazon S3 buckets, or Amazon Data Firehose streams, and deliver to multiple destinations simultaneously. This significantly enhances observability, enabling richer analytics (e.g., Parquet to S3 for Athena) and flexible alerting. Security-wise, improved logging is paramount for forensic analysis, threat detection, and auditing, supporting compliance requirements like PCI DSS and HIPAA. Architecturally, it facilitates centralized log management and integration with SIEM/SOAR solutions. *Terraform Support*: Pending provider release for specific arguments related to configurable execution log destinations and larger log events (use AWS CLI/SDK or custom provisioner in the interim).

Amazon MQ now supports RabbitMQ version 4.3, introducing significant quorum queue feature enhancements such as compaction, increased priority levels, native delayed retries, and graceful consumer timeouts. This update improves message durability, performance, and operational flexibility for message brokers. From a security standpoint, enhanced durability and retry mechanisms reduce the risk of message loss and improve data integrity. Architecturally, these features provide more robust and efficient messaging patterns for building resilient distributed applications. *Terraform Support*: Supported in Terraform AWS Provider: aws_mq_broker (via the engine_version argument) in v5.35.0+ up to >= 5.49.0.

Media Services Innovation

AWS Elemental MediaTailor now offers 'Yield Optimization,' automatically filling previously unused ad inventory in livestreams with Amazon Ads demand during server-side ad insertion (SSAI). This feature provides seamless ad stitching, scalability for large events, and brand safety features (price floors, category rules), configurable via API or Console. Security benefits include brand safety controls that protect content integrity and advertiser reputation. Architecturally, it optimizes monetization for live streaming without requiring additional infrastructure, improving ROI. *Terraform Support*: Pending provider release (feature-specific parameters not yet available) (use AWS CLI/SDK or custom provisioner in the interim).

AWS Elemental Inference now generates contextual metadata from live video streams in real time using AI. This includes IAB content taxonomy categories, GARM brand suitability signals, detected objects/actions, and scene descriptions. This metadata can be embedded into SCTE-35 ad markers by MediaLive, enabling context-aware ad decision-making via MediaTailor. Security-wise, GARM signals are crucial for ensuring brand suitability and protecting brand reputation. Architecturally, this enables advanced monetization, content enrichment, and content discovery workflows for live video. *Terraform Support*: Supported in Terraform AWS Provider: aws_medialive_channel (for integrating with MediaLive channels) in v5.45.0+ up to v5.70.0.

AWS Elemental MediaLive now supports A/B forensic watermarking, enabling content owners to trace the source of unauthorized redistribution of live video content. A single MediaLive channel produces two synchronized output variants, each with a distinct, visually transparent watermark that persists through re-encoding. This follows the DASH Industry Forum (DASH-IF) specification and is configurable on CMAF Ingest output groups. This is a critical security feature for content protection and intellectual property rights enforcement, deterring piracy. Architecturally, it provides a robust and scalable mechanism for identifying content leaks. *Terraform Support*: Supported in Terraform AWS Provider: aws_medialive_channel (for configuring output group settings) in v5.68.0+ up to >= 5.72.0.

AWS Elemental MediaTailor now enables Low-Latency HLS (LL-HLS) ad insertion through HLS Interstitials. This mechanism delivers partial segments and keeps playlist requests open, significantly reducing live stream latency. By referencing ads as separate playlists, it optimizes CDN performance and scalability by ensuring every viewer receives the same cacheable media playlist. Security benefits include maintaining stream integrity during ad insertion, reducing potential attack surfaces. Architecturally, this is vital for delivering high-quality, low-latency live streaming experiences with effective monetization. *Terraform Support*: Supported in Terraform AWS Provider: aws_mediatailor_playback_configuration (for ad and HLS configuration) in v2.1.0+ up to ~> 5.0.

AWS Elemental MediaPackage now offers Dynamic Multiview, a server-side capability that composes multiple live video sources into viewer-selected tiled layouts on demand. It delivers these as standard HLS and DASH streams, playable on most modern devices without custom player development. Operating entirely in the compressed domain, it combines individually encoded sources without re-encoding. Security benefits include support for DRM encryption. Architecturally, this enables highly personalized and interactive live viewing experiences efficiently, reducing encoding costs and simplifying player development. *Terraform Support*: Supported in Terraform AWS Provider: aws_mediapackage_origin_endpoint (for HLS or DASH package configurations) in v5.26.0+ up to >= 5.26.0.

Databases, Monitoring & Analytics

Amazon Redshift RG instances, powered by AWS Graviton processors, are now available in the AWS Europe (Zurich) Region. These instances offer up to 2.4x faster performance than previous-generation RA3 instances at a 30% lower price per vCPU. They include Redshift's custom-built vectorized data lake query engine for Apache Iceberg and Parquet data. Security benefits include enhanced hardware-level security features of Graviton processors. Architecturally, this provides high-performance, cost-effective data warehousing and lakehouse capabilities, crucial for data residency requirements in Switzerland. *Terraform Support*: Supported in Terraform AWS Provider: aws_redshift_cluster (via the node_type argument) in v5.15.0+ up to v5.47.0.

CloudWatch Network Monitoring synthetic monitors can now determine if network performance issues on paths crossing an AWS Transit Gateway inter-Region peering connection are caused by the AWS network. The Network Health Indicator (NHI) now covers these paths, publishing health data to your CloudWatch account for dashboards and alarms. This significantly helps network operators and application developers isolate the source of degradation. From a security perspective, improved visibility into network health aids in identifying potential attack vectors or performance issues that could be exploited. Architecturally, it's crucial for diagnosing inter-region connectivity in complex hybrid/multi-region setups, reducing Mean Time To Resolution (MTTR). *Terraform Support*: Supported in Terraform AWS Provider: aws_synthetics_canary (for configuring synthetic monitors) in v5.35.0+ up to >= 5.35.0.

You can now build full-stack search and AI applications rapidly using Amazon OpenSearch Serverless on v0 by Vercel, leveraging natural language prompts for full-text and vector search in RAG workloads. v0 automates the provisioning of an OpenSearch Serverless collection, data indexing, endpoint configuration, and environment variables, following recommended patterns. Security benefits include leveraging OpenSearch Serverless's inherent security model (VPC access, encryption at rest/in transit). Architecturally, this integration simplifies and accelerates the deployment of AI-powered search applications, fostering innovation. *Terraform Support*: Supported in Terraform AWS Provider: aws_opensearchserverless_collection (for managing the underlying OpenSearch Serverless collection) in v4.60.0+ up to v5.72.0.

Announcements in this summary