AWS News - 2026-09-29

AWS GovCloud (US) now offers Claude Sonnet 5.5, a smarter and more efficient Sonnet that delivers a clear leap forward on coding and knowledge work with at a lower cost per task for most work ag faster speed.

Claude Sonnet 5.5 is a smarter, more efficient Sonnet and a step up from Sonnet 5, making it a natural upgrade for teams already building on Sonnet. It's stronger for coding, completing well-scoped tasks as part of a larger coding strategy such as building and fixing features with Claude in the same session or verifying output against requirements. And it produces ready-to-share knowledge work, from 1-pagers, diagrams, and summary slides to document edits and spreadsheet cleanup.

Amazon Bedrock gives you Sonnet's capabilities while keeping your data within AWS infrastructure with regional data residency and provides access through a unified service with AWS-managed features like Guardrails and Knowledge Bases. To learn more, see the Amazon Bedrock documentation and regional availability. 

Amazon Bedrock now supports SpaceXAI Grok 4.7, a frontier model built for coding, agentic tasks, and knowledge work, with US Geo and Global cross-Region inference.  

Grok 4.7 builds on Grok 4.6 with better mixed-document handling, more dependable repo-scale coding with planning and error recovery, and enhanced browser-use agents for form fills and portal navigation.  

Get started with Grok 4.7 in the Amazon Bedrock console or programmatically through supported Amazon Bedrock APIs. For information about supported AWS Regions, endpoints, APIs, features, inference profiles and pricing, see the Amazon Bedrock documentation. To learn more, read the launch blog post.

 

Amazon Rekognition Face Liveness now returns Feedback Codes that help customers understand why a liveness check received a low score and how users could retry successfully. The codes are returned in the GetFaceLivenessSessionResults API response, and they identify the specific conditions that contributed to the low score, such as poor lighting, face obstruction, closed eyes, or low video quality. Customers can surface these codes to their end users who can then make targeted corrections and retry with confidence. A single session can return multiple Feedback Codes at once, allowing users to resolve all contributing issues in one retry. This feature is especially useful in identity verification onboarding, re-authentication flows, and accessibility scenarios.

Amazon Rekognition Face Liveness verifies that the person in front of the camera is a real, live human (instead of a photo, video replay, mask, or deepfake) through a short, guided selfie-video check. It returns a confidence score, a reference image for face matching, and session audit images. 

Rekognition Face Liveness is available in US East (N. Virginia), US West (Oregon), Europe (Ireland), Asia Pacific (Tokyo, Mumbai, Malaysia, Thailand), and South America (São Paulo). To learn more, visit the Amazon Rekognition Face Liveness API documentation and Face Liveness FAQ.

Amazon EC2 Future-dated Capacity Reservations let you secure capacity up to 120 days in advance by specifying the capacity you need, the start date, and a commitment duration. If your plans change before delivery, you can now postpone your start date to match your new timeline.

When you postpone your start date, EC2 presents a quote that shows the new start date and the required commitment duration. The commitment duration can increase, depending on when you postpone the start date. After you review and accept the quote, EC2 updates your reservation with the new start date.

This feature is available to all Future-dated Capacity Reservation customers. Refer to the AWS Capabilities by Region website for the feature's regional availability. For more details, refer to the Capacity Reservation user guide.

Amazon ElastiCache Serverless for Valkey now supports public endpoints, letting you connect to your cache directly from a laptop, a serverless function, or any application running outside AWS, without setting up a VPN, a bastion host, or an SSH tunnel.

With a public endpoint, ElastiCache Serverless gives you a fully managed cache reachable over the internet, with no VPC to configure and no infrastructure to provision, and you can create one in under a minute. Use it to rapidly prototype, connect AI coding tools and agents straight to your cache, or add caching to workloads that can't reach a VPC. Every connection uses IAM authentication over TLS 1.3, so there's no password to store or rotate. Connect using Valkey GLIDE 2.2 or later, which has built-in IAM support, or another Valkey client paired with the Developer Toolkit for ElastiCache, an open-source library that generates and refreshes IAM authentication tokens for you.

Public endpoints for ElastiCache Serverless are available in all commercial AWS Regions and the China Regions. There is no additional charge for using public endpoints beyond standard ElastiCache Serverless pricing. To get started, create a Valkey 9.0 or later serverless cache with a public endpoint using the AWS Management Console, AWS SDK, or AWS CLI. To learn more about public endpoints, see Create a Valkey serverless cache with a public endpoint.

AWS Systems Manager now lets you share your Systems Manager Documents (SSM Documents) with an entire AWS organization or with specific organizational units (OUs) using AWS Resource Access Manager (AWS RAM). Previously, you could only share a document publicly or with a list of individual account IDs. You can now easily share a document with your organization or OUs, and Systems Manager automatically keeps that sharing up to date as accounts are added to or removed from your organization or OU, so you no longer need to track and update accounts manually.

To share a document, you create an AWS RAM resource share, add the SSM Documents you want to share, and select the organizations or OUs to share them with. Because sharing is managed through AWS RAM and resource-based policies, access is granted through standard AWS authorization. When you share a document with an account outside your organization, that account receives a resource share invitation and gains access only after it accepts, giving both the document owner and the consumer more control over shared access.

This capability is available in the Systems Manager console, the AWS Command Line Interface (AWS CLI), and the AWS SDKs, in all AWS Regions where AWS Systems Manager is available. There is no additional charge to share documents through AWS RAM. To get started, open the Documents page in the Systems Manager console, or see the AWS Systems Manager User Guide at https://docs.aws.amazon.com/systems-manager/latest/userguide/documents-ssm-sharing.html.

AWS DataSync now supports shared Virtual Private Clouds (VPCs). You can create DataSync agents and run transfer tasks using subnets shared across AWS accounts with AWS Resource Access Manager (RAM). It allows you to transfer data privately over a shared subnet and VPC endpoint managed in a central account, rather than one per account.

Customers that centralize their networking previously had to create a separate DataSync VPC endpoint in every account that connected privately through AWS PrivateLink. Each endpoint consumed IP addresses and added operational overhead to maintain across accounts. With shared VPC support, a single endpoint in the account that owns the VPC serves every account the subnet is shared with, conserving IP address space and removing the need for a per-account endpoint. This launch also helps within a single account setup. You can now use one VPC endpoint across multiple subnets, removing the earlier requirement for a matching VPC endpoint in each subnet.

Shared VPC is supported for both Enhanced mode and Basic mode agent-based tasks. To get started, create a DataSync agent using the console or the CreateAgent API and specify a subnet shared with your account via AWS RAM. AWS DataSync support for Shared VPCs is available in all AWS Regions, except AWS Secret Regions, where AWS DataSync is offered. To learn more, visit the AWS DataSync feature documentation.

AWS Transfer Family now lets web app users download multiple files and folders at once. End users can select any combination of files and folders and download them together in a single action. The download is delivered as a zip archive that preserves the original folder structure when unzipped.

You can get started by selecting files and folders in a location and choosing Download from the action menu. Previously, end users could download only one file at a time, and folders could not be downloaded at all. The web app shows the files to be downloaded and real-time progress with a success or failed status for each one.

Multiple-file download is available in Google Chrome, Mozilla Firefox, and Chromium-based browsers such as Microsoft Edge. It is not currently supported in Safari, where end users can still download files one at a time. This capability is available in all AWS Regions where Transfer Family Web Apps is offered. To learn more, visit the Transfer Family web apps User Guide.

Amazon CloudWatch Logs now automatically indexes the fields you query frequently, speeding up CloudWatch Logs Insights queries without requiring any manual setup. Previously, you had to manually select the fields to index based on your query patterns and usage, requiring extra work for teams whose query patterns change over time.

CloudWatch Logs identifies the fields in your queries that you use for filtering with the = and IN operations and indexes them automatically, enabling queries to run faster and scan less data. Fields that are automatically indexed don't count toward your 20-field-per-log-group limit and are retained for 30 days. The list of automatically indexed fields is updated as your query patterns change. To keep a field indexed permanently, promote it to a field index policy using the console or API.

Auto-indexing is available in all AWS Regions where CloudWatch Logs field indexing is supported, at no additional cost. To learn more, see the Amazon CloudWatch Logs User Guide.

Amazon CloudWatch Logs Insights now lets you estimate the volume of log data, in bytes, that a query would scan over your selected log groups and time range - without running the query. This helps you adjust your log group selection, time range, and filters before running the query.

In the CloudWatch console, the estimate appears automatically in the query editor when you change the log group selection, time range, or query text. When using the AWS CLI or API, append the estimate command to your query to request it explicitly. Queries that use the estimate command incur no CloudWatch Logs Insights query charges.

The estimate command is available in all AWS commercial regions. To learn more, see the Amazon CloudWatch Logs documentation.

If there’s one theme that defined last week, it’s choice. The frontier models keep arriving, and the interesting question is no longer just “how smart is it?” but “which model fits this step, at this cost, at this latency?” That’s exactly what landed on Amazon Bedrock over the past few days: GPT-6 Sol and GPT-6 […]

Claude Opus 5.5 が 2026 年 9 月 28 日より Kiro IDE、CLI、Crew、Web のすべてで利用可能になりました。Anthropic のこれまでで最も高性能な Opus モデルであり、ほとんどの作業において Fable 5.1 と同等の性能を、Opus 5 より 40% 低いコストで提供します。

送配電事業（Transmission & Distribution Business）に携わる皆様は、これまでにない変化の波を肌で感じていらっしゃるのではないでしょうか。熟練技術者の高齢化や定年退職が加速し暗黙知が失われる一方で、再エネ大量導入に伴う系統の複雑化、設備の老朽化と新規投資の削減圧力、そして量子コンピュータ時代を見据えたサイバー脅威の高度化 ― 複合的な課題が同時に押し寄せ、従来の「人海戦術＋経験則」だけでは対処しきれない時代が到来しています。

はじめに 本記事は、HR データ統合基盤を題材に、共通オントロジーの設計の考え方と、取込・質問応答・推論への使 […]

On Saturday, October 24, 2026, we will conduct an exercise demonstrating that the AWS European Sovereign Cloud can operate without depending on any infrastructure outside of the European Union (EU). For several hours, the AWS European Sovereign Cloud will operate without a connection to the AWS Global Network backbone. The backbone is the private network […]

Learn how to build AI interfaces that automatically adapt to your agents' variable outputs, using the AG-UI protocol for dynamic UI generation, the Strands Agents SDK swarm pattern for explainable multi-agent collaboration, and Amazon Nova Act to integrate legacy systems that lack APIs.

Learn how to operationalize Amazon Textract Custom Queries adapters for production: infrastructure as code with AWS CloudFormation and Terraform, a cross-account adapter promotion process, a pre-classification routing pattern for multiple form versions, and production security controls such as VPC endpoints, encryption, and least-privilege IAM.

Learn an agent-driven approach to synthetic monitoring using Amazon Nova Act and Amazon Bedrock AgentCore. The post covers the architecture and patterns for resilient, managed user-journey validation that moves beyond brittle UI scripts, with a complete sample implementation.

Deploy two generative media models from one AWS vLLM-Omni Deep Learning Container on Amazon SageMaker AI. Generate an image with FLUX.2-klein through real-time inference, then animate it into video with Wan2.1-VACE through asynchronous inference, and retrieve the MP4 from Amazon S3.

Deploy a text-to-speech model on Amazon SageMaker AI with the AWS vLLM-Omni Deep Learning Container and stream generated speech over a persistent bidirectional connection. This Part 1 tutorial deploys Qwen3-TTS and streams speech through a Gradio application.

Claude Sonnet 5.5 is now available on Amazon Bedrock and Claude Platform on AWS. It's a smarter, more efficient Sonnet model for focused coding and knowledge work, with a lower cost per task at faster speed. This post covers what's new, when to choose Sonnet, and how to get started.

xAI's Grok 4.7 is now available on Amazon Bedrock: a frontier model for coding, long-running agents, and knowledge work. It offers a 500K token context window and four configurable reasoning effort levels, reachable through the Responses, Chat Completions, and Converse APIs.

AWS Lambda now supports scalable network bandwidth for functions with 2,048 MB of memory or more running outside a VPC. Sustained throughput scales from 625 Mbps up to 3,000 Mbps at 10,240 MB, reducing execution times and per-invocation costs for latency-sensitive data processing workloads.

Amazon EventBridge relaunched the Custom event bus so a platform team can share one governed bus across the organization while application teams publish and subscribe from their own accounts. See how retention, ordering, open formats, transformation, and direct target delivery change what one bus can carry.