New updates on AWS GovCloud (US) | Issue #9 (2026)
September 16 - 30, 2026 | New updates on AWS GovCloud (US): This issue features insights on how a U.S. Army R&D team is accelerating software delivery with AI-powered agentic engineering on AWS GovCloud (US), and getting started with Anthropic Claude Opus 5.5 and Claude Sonnet 5.5 on Amazon Bedrock in AWS GovCloud (US), plus the latest features and services.
Executive Insights
U.S. Army R&D Accelerates Software Delivery with AI-Powered Agentic Engineering on AWS GovCloud (US)
A U.S. Army R&D team and AWS built a secure agentic engineering solution on AWS GovCloud (US) that turns 12-month authorization cycles into 3-month MVP deployments, all within DoD CSP SRG IL 2/4/5 compliance. A secure Landing Zone Accelerator implemented with Zero Trust architecture lets distributed government and industry teams work together without dedicated circuits or government-issued hardware.
AI-powered agentic engineering automates scanning for compliance with DISA Security Technical Implementation Guides, reducing the process from 3 weeks to 2-3 hours. New programs inherit the solution’s Authority to Operate (ATO) instead of building their own, with built-in continuous security controls and automated compliance monitoring. The result is a scalable, replicable model that accelerates defense software delivery and reduces onboarding from months to weeks.
Other R&D organizations interested in modernizing their software development capabilities can explore how the secure coding environment on AWS GovCloud (US) enables more secure, agile DevSecOps at scale. To learn more, read the full case study on how the Army R&D team uses agentic engineering on AWS GovCloud (US) or explore the Landing Zone Accelerator on AWS.
Solutions Insights
Getting Started with Anthropic Claude Opus 5.5 and Claude Sonnet 5.5 in AWS GovCloud (US)
Here’s what practitioners need to know to get started with the new Claude 5.5 models in AWS GovCloud (US).
Anthropic Claude Opus 5.5 and Claude Sonnet 5.5 are now available on Amazon Bedrock in both AWS GovCloud (US-East) and AWS GovCloud (US-West) Regions. Both models are accessible through the bedrock-runtime endpoint in AWS GovCloud (US-East) and AWS GovCloud (US-West) Regions, and through the bedrock-mantle endpoint in the AWS GovCloud (US-West) Region, and support the Messages, Responses, Chat Completions, Converse, and Invoke APIs. Both share a 1M-token context window, 128K max output tokens, and multimodal support for text, image, audio, speech, and video input and output.
Claude Opus 5.5 improves on Opus 5 for long-running coding and knowledge work, completing tasks with fewer tokens at a lower price per token, with cheaper cache reads compounding the cost reduction. Adaptive thinking lets Claude dynamically decide when and how much to reason based on the complexity of each request, instead of requiring a fixed thinking token budget. In adaptive mode, Claude evaluates the input, decides if chain-of-thought reasoning is needed, and calibrates its depth accordingly. The default effort level is medium, with selectable options of low, high, xhigh, and max to let teams tune performance against latency and cost. Opus 5.5 also supports Compaction, a capability that summarizes prior conversation context to stay within token limits without losing critical information during extended sessions. Opus 5.5 reports clearly on completed work, findings, and next steps, making it well suited for sustained, multi-step workloads. Amazon Bedrock provides both models with zero data retention (ZDR) support by default.
Claude Sonnet 5.5 improves on Sonnet 5 with higher performance on coding and knowledge work at lower cost and faster response times. Sonnet 5.5 handles well-scoped coding tasks such as building and fixing features in the same session and verifying output against requirements. It also produces polished knowledge work including summaries, diagrams, slides, document edits, and spreadsheet cleanup. Adaptive thinking is on by default with configurable reasoning effort levels (low, medium, high, xhigh, max, default: high). Teams already using previous Sonnet models on Amazon Bedrock can upgrade to Sonnet 5.5 to get faster responses, stronger coding and reasoning, and lower cost per task. Update to model ID anthropic.claude-sonnet-5-5 and enable access in the Amazon Bedrock console to start.
Learn more: Opus 5.5 Documentation | Sonnet 5.5 Documentation | Opus 5.5 What’s New Post | Sonnet 5.5 What’s New Post
Service and Feature Releases
Source: What’s New Posts, Sept 16-30, 2026.
Service Releases
None
Feature Releases
Helpful Links
Did you find what you were looking for today?
Let us know so we can improve the quality of the content on our pages