AWS News - 2026-10-24

Amazon Connect Customer Global Resiliency now provides customers a mechanism to deliver long, persistent chat experiences across linked AWS Regions. With this capability, end users can resume previous conversations with full context, metadata, and transcripts carried over across Regions, eliminating the need for end users to repeat information and allowing agents to provide personalized service with access to the entire conversation history across AWS Regions.

For example, an end user can start a chat in US-East (N. Virginia) and later come back to it in the linked US-West (Oregon) Region, continuing the same conversation with an agent with full conversation history and chat attachment. To adopt this capability, onboard to Global routing.

This feature is available in all AWS Regions where Amazon Connect Customer Global Resiliency is supported. To learn more, visit Amazon Connect Customer Admin Guide.

Web Search on AWS Bedrock now supports two new controls that give you more precision over the results that ground a model's response: domain filtering and user location. Both are configured directly on the web_search tool in your existing OpenAI Responses API request.

With domain filtering, you can restrict Web Search results to a set of domains you choose. Using the filters parameter, you specify up to 100 allowed_domains to limit results to trusted sources, or up to 100 blocked_domains to exclude sources you don't want. This is useful when you want responses grounded only in specific sites, or when you need to keep certain domains out of results.

With user location, you can optionally specify location parameters so that Web Search can return results relevant to that geography. This helps with location-dependent questions, such as regional availability or local services.

Domain filtering and user location are available in all AWS Regions where Web Search is supported: US East (N. Virginia), US East (Ohio), US West (Oregon), and AWS GovCloud (US-West). To get started, see the Web Search section in the AWS Bedrock User Guide. For pricing, see the AWS Bedrock pricing page.

Web Search on Amazon Bedrock now supports two new controls that give you more precision over the results that ground a model's response: domain filtering and user location. Both are configured directly on the web_search tool in your existing OpenAI Responses API request.

With domain filtering, you can restrict Web Search results to a set of domains you choose. Using the filters parameter, you specify up to 100 allowed_domains to limit results to trusted sources, or up to 100 blocked_domains to exclude sources you don't want. This is useful when you want responses grounded only in specific sites, or when you need to keep certain domains out of results.

With user location, you can optionally specify location parameters so that Web Search can return results relevant to that geography. This helps with location-dependent questions, such as regional availability or local services.

Domain filtering and user location are available in all AWS Regions where Web Search is supported: US East (N. Virginia), US East (Ohio), US West (Oregon), and AWS GovCloud (US-West). To get started, see the Web Search section in the Amazon Bedrock User Guide. For pricing, see the Amazon Bedrock pricing page.

Bulletin ID: 2026-148-AWS
Scope: AWS
Content Type: Important (requires attention)
Publication Date: 10/23/2026 10:30 AM PDT

Description:

Amazon Web Services Multi Model Server (MMS) is an open source tool for serving machine learning models for inference. It was published by AWS on GitHub, PyPI, and Docker Hub, and reached end of life; its source repository has been archived and read-only since November 20, 2025.

CVE-2026-115812: An issue in model archive extraction could allow a crafted model archive to write files outside the intended directory, potentially resulting in code execution with the privileges of the MMS process.

CVE-2026-115813: An issue in the model download path could allow the MMS management interface to be directed to make requests to unintended network destinations.

Impacted versions: <=1.1.11 (all released versions)

Please refer to the article below for the most up-to-date and complete information related to this AWS Security Bulletin.

Read this article

Bulletin ID: 2026-149-AWS
Scope: AWS
Content Type: Important (requires attention)
Publication Date: 10/23/2026 12:00 PM PDT

Description:

AWS MCP Server is a managed server that gives AI agents access to AWS through the Model Context Protocol (MCP), including a run_script tool that runs Python scripts in a sandboxed environment.

We identified CVE-2026-115845, an issue in the Python code validator of the run_script tool in AWS MCP Server. An incomplete list of disallowed inputs might allow an authenticated remote user to execute arbitrary code within their own isolated session environment via a crafted script.

Impacted versions: N/A (managed service, exclusively hosted)

Please refer to the article below for the most up-to-date and complete information related to this AWS Security Bulletin.

Read this article

Bayer Crop Science built the Agentic Decision Engine (ADE), a multi-agent platform on Amazon Bedrock AgentCore that lets breeders, analysts, and researchers ask natural language questions about their analytics platform. This post covers the ADE architecture, five design decisions, and how the team prepared it for production with AWS Unified Operations.

Amazon Quick Sight layer maps now support Point and LineString geometry from the same GeoJSON workflow, so you can plot store locations and the routes between them. This post covers the dataset and GeoJSON structure layer maps require, when to choose points or lines, and techniques for viewport control, layering geometries, and dynamic GeoJSON updates.

Updating a SageMaker AI endpoint that has Application Auto Scaling registered makes AWS CloudFormation and AWS CDK deployments fail. This post shows how to build a reusable AWS CDK construct with CloudFormation custom resources that deregisters and re-registers auto scaling automatically, so you can update the endpoint and its scaling in a single deployment.

Multi-Turn Reinforcement Learning (MTRL) on Amazon SageMaker AI trains models on the outcome of a whole multi-step task rather than labeled steps. This hands-on guide walks the full pipeline using Wordle: deploying the agent and reward on Amazon Bedrock AgentCore, preparing the dataset, training, and evaluating at serving settings.