All news with #aws tag

AWS Organizations Adds Full IAM Policy Language to SCPs

🔐 AWS Organizations now supports the full IAM policy language for service control policies (SCPs), allowing administrators to use conditions, individual resource ARNs, and the NotAction element with Allow statements. You can also apply wildcards at the beginning or middle of Action strings and use the NotResource element for finer scoping. These enhancements let teams create more concise and precise organizational guardrails to enforce least-privilege across accounts. The change is backward compatible and available in all AWS commercial and AWS GovCloud (US) Regions.

AWS Neuron SDK 2.26 Adds Trn2, PyTorch 2.8, JAX 0.6.2

🚀 AWS has released Neuron SDK 2.26.0 as generally available, delivering framework and runtime improvements for Inferentia and Trainium-based instances. The update adds support for PyTorch 2.8 and JAX 0.6.2, enhances inference on Trainium2 (Trn2) instances, and enables deployment of models such as FLUX.1-dev and beta Llama 4 Scout/Maverick. It also introduces expert parallelism (beta) for MoE models, new Neuron Kernel Interface APIs, and an improved Neuron Profiler with system profile grouping for distributed workloads.

AWS SiteWise MCP Server Accelerates Industrial Modeling

⚙️ AWS published a Model Context Protocol (MCP) server for AWS IoT SiteWise in the AWS Labs open-source MCP repository to simplify industrial data modeling. The server embeds domain validation and automated modeling, applying correct units, data types, and quality indicators so models are production-ready. It maintains compatibility with existing SiteWise tools and APIs while adding conversational interfaces to streamline model authoring, asset onboarding, and downstream analytics enablement.

Source-of-Truth Authorization for RAG Knowledge Bases

🔒 This post presents an architecture to enforce strong, source-of-truth authorization for Retrieval-Augmented Generation (RAG) knowledge bases using Amazon S3 Access Grants with Amazon Bedrock. It explains why vector DB metadata filtering is insufficient—permission changes can be delayed and complex identity memberships are hard to represent—and recommends validating permissions at the data source before returning chunks to an LLM. The blog includes a practical Python walkthrough for exchanging identity tokens, retrieving caller grant scopes, filtering returned chunks, and logging withheld items to reduce the risk of sensitive data leaking into LLM prompts.

Amazon VPC Reachability and Network Access Analyzer Expand

🛰️ Amazon has expanded VPC Reachability Analyzer and VPC Network Access Analyzer to seven additional regions — New Zealand, Hyderabad, Melbourne, Taipei, Calgary, Tel Aviv, and Mexico Central. Reachability Analyzer diagnoses network reachability between source and destination resources, while Network Access Analyzer identifies unintended access paths that may bypass security controls. This regional launch improves troubleshooting, compliance checks, and multi-account network visibility; pricing and documentation are available through AWS resources.

AWS Outposts expand to Canada (Central) and N California

📢 Second-generation AWS Outposts racks are now supported in the AWS Canada (Central) and US West (N. California) Regions. Outposts racks extend AWS infrastructure, services, APIs, and tools to on-premises data centers or colocation spaces, providing a consistent hybrid experience. Customers can order racks connected to these Regions to optimize for latency and data residency, run low-latency workloads locally, and maintain centralized management in their home Region.

AWS Kinesis Data Streams Adds IPv6 and FIPS in GovCloud

🌐 Amazon Web Services announced that Kinesis Data Streams now supports API requests over IPv6 in the AWS GovCloud (US) Regions, with optional dual-stack (IPv4/IPv6) public and VPC endpoints. The new endpoints have been validated under FIPS 140-3, enabling FIPS-compliant encryption for customers contracting with the US federal government. IPv6 support reduces address overlap and simplifies connectivity for devices and networks already using IPv6. This capability is available in all Regions where Kinesis Data Streams operates, including GovCloud and China Regions.

Amazon Q Developer CLI Adds Remote MCP Server Support

🔒 Amazon Q Developer CLI now supports remote MCP servers to centralize tool integrations and OAuth-based authentication, enhancing scalability and security in development workflows. Administrators specify HTTP transport, the authentication URL, and optional headers in agent configuration or mcp.json. Upon successful OAuth authentication, the CLI enumerates tools on the MCP server and exposes them to the agent. This capability is available in both the CLI and the Amazon Q Developer IDE plugins.

Stability AI Image Services Now Available in Amazon Bedrock

🖼️ Amazon Bedrock now includes Stability AI Image Services, a suite of nine specialized image-editing tools available via the Bedrock API. The offering splits into Edit tools (Remove Background, Erase Object, Search and Replace, Search and Recolor, Inpaint) and Control tools (Structure, Sketch, Style Guide, Style Transfer). It is currently supported in US West (Oregon), US East (N. Virginia), and US East (Ohio), and is intended to accelerate professional creative workflows with granular edit control.

Amazon Bedrock Adds Four Qwen3 Open-Weight Models Now

🤖 Amazon Web Services added four Qwen3 open-weight foundation models to Amazon Bedrock as fully managed, serverless offerings. The lineup—Qwen3-Coder-480B-A35B-Instruct, Qwen3-Coder-30B-A3B-Instruct, Qwen3-235B-A22B-Instruct-2507, and Qwen3-32B—covers both dense and Mixture-of-Experts (MoE) architectures. The coder variants specialize in agentic coding, function calling, and tool use, while the 235B and 32B models provide general reasoning and efficient dense computation. These models are available now across multiple AWS regions, enabling developers to build advanced AI applications without managing infrastructure.

OpenAI Open-Weight Models Now in Eight More AWS Regions

🚀 AWS has expanded availability of OpenAI open weight models on Amazon Bedrock to eight additional regions. The update adds US East (N. Virginia), Asia Pacific (Tokyo), Europe (Stockholm), Asia Pacific (Mumbai), Europe (Ireland), South America (São Paulo), Europe (London), and Europe (Milan) to the previously supported US West (Oregon). This broader regional coverage reduces network latency, helps meet data residency preferences, and makes it easier for customers to deploy AI-powered applications closer to their users. Customers can access the models through the Amazon Bedrock console and supporting documentation to get started.

Step Functions: Data Sources and Metrics for Distributed Map

⚙️ AWS Step Functions now expands Distributed Map input sources and adds visibility metrics. Distributed Map can now iterate S3 objects via S3ListObjectsV2, read AWS Athena data manifests and Parquet files directly, and extract arrays from JSON stored in S3 or passed as state input. New observability metrics — Approximate Open Map Runs Count, Open Map Run Limit, and Approximate Map Runs Backlog Size — provide operational insight. These features are available in all commercial AWS Regions; enable Distributed Map mode in the Step Functions console and consult the developer guide for examples.

DeepSeek-V3.1 Available as Fully Managed in Bedrock

🔍 DeepSeek-V3.1 is now available as a fully managed foundation model in Amazon Bedrock, offering an open-weight option designed for enterprise deployment. The model supports a selectable 'thinking' mode for step-by-step analysis and a faster non-thinking mode for quicker replies, with improved multilingual accuracy and reduced hallucinations. Enhanced tool-calling, transparent reasoning, and strong coding and analytical performance make it well suited for building AI agents, automating workflows, and tackling complex technical tasks. DeepSeek-V3.1 is available in US West (Oregon), Asia Pacific (Tokyo, Mumbai), and Europe (London, Stockholm).

AWS Bedrock Adds OpenAI Open‑Weight Models in Eight Regions

🚀 AWS has expanded availability of OpenAI open weight models on AWS Bedrock to eight additional AWS Regions worldwide. The update brings the models to US East (N. Virginia), Asia Pacific (Tokyo, Mumbai), Europe (Stockholm, Ireland, London, Milan) and South America (São Paulo), alongside existing US West (Oregon) support. This broader footprint aims to lower latency, improve model performance and help customers meet data residency requirements. To get started, use the Amazon Bedrock console or consult the documentation.

Amazon Lex: confirmation and currency slots in 10 languages

🤖 Amazon Lex now supports built-in confirmation and currency slot types in 10 additional languages: Portuguese, Catalan, French, Italian, German, Spanish, Mandarin, Cantonese, Japanese, and Korean. These built-in slots normalize varied user phrasing—mapping acknowledgements to 'Yes', 'No', 'Don't know', or 'Maybe' and converting currency expressions into structured formats such as 'USD 1.00'—to simplify multi-lingual conversational flows. The feature is available in all commercial AWS Regions where Amazon Lex operates and can improve chatbots and contact-center interactions.

Amazon SageMaker HyperPod Adds Managed Karpenter Autoscaling

🛠️ Amazon SageMaker HyperPod now supports managed node autoscaling using Karpenter, enabling automated cluster scaling for both inference and training workloads. This managed capability removes the operational burden of installing and maintaining autoscaling infrastructure while providing integrated resilience and fault tolerance. Customers gain just-in-time GPU provisioning, scale-to-zero during low demand, workload-aware instance selection, and cost reductions through intelligent consolidation.

Amazon EVS Adds HCX Migration Over Public Internet

🌐 Amazon EVS now supports VMware HCX migrations over the public internet using Elastic IP Addresses (EIPs) to provide stable endpoints and faster setup. This option supplements existing private connectivity methods such as AWS Direct Connect and VPN, enabling secure layer‑2 network stretch and workload migration when private links are unavailable. Public HCX connectivity is available in all AWS Regions where EVS is offered and can be a cost‑effective alternative for workloads that do not require private connection performance.

AWS Step Functions Adds IPv6 Dual-Stack Endpoint Support

🌐 AWS Step Functions now supports IPv6 via new dual-stack IPv4/IPv6 endpoints, enabling customers to send IPv6 traffic directly to the service. The enhancement preserves backwards compatibility with existing IPv4 endpoints and enables PrivateLink interface VPC endpoint connectivity so workloads can access Step Functions privately without traversing the public internet. IPv6 support is generally available in several US commercial and GovCloud regions.

Amazon OpenSearch Serverless Adds Disk-Optimized Vectors

🔍 Amazon has added disk-optimized vector storage to OpenSearch Serverless, offering a lower-cost alternative to memory-optimized vectors while maintaining equivalent accuracy and recall. The disk-optimized option may introduce slightly higher latency, so it is best suited for semantic search, recommendation systems, and other AI search scenarios that do not require sub-millisecond responses. As a fully managed service, OpenSearch Serverless continues to automatically scale compute capacity (measured in OCUs) to match workload demands.

AWS Network Firewall: SNI Session Holding for TLS Guide

🔒 AWS Network Firewall now offers SNI session holding to strengthen TLS inspection by validating the TLS SNI before initiating an outbound TCP connection. When enabled, the firewall holds TCP/TLS establishment until it receives the ClientHello SNI and evaluates it against Suricata-based TLS inspection rules, preventing any contact with disallowed endpoints. Administrators can enable this option in a TLS inspection configuration via the AWS Management Console, AWS CLI, or AWS SDK; it’s available in Regions including GovCloud and China and is billed as part of TLS advanced inspection.