AI news for builders and product teamsUpdated Nov 17, 2026, 13:01 UTC
AWS Machine Learning Blog
First-party releases and research from AWS Machine Learning Blog. Headlines and excerpts link to the original articles.
Latest stories
Newest first
Orca Security's Discovery AI, built on Amazon Bedrock, converts plain-English security questions into Sonar queries, cutting query costs 90 percent, input tokens 93 percent and latency 75 percent versus its initial full-schema prompt baseline.

ShipStation Global built Deckhand, an AI IT help desk agent for Slack on Amazon Bedrock AgentCore that resolved 97.5 percent of incoming channel queries and earned 96 percent positive feedback in its first three months.

SpaceXAI's Grok 4.7 is now available on Amazon Bedrock, where a tool-calling agent evaluation found that running the model at its xhigh reasoning effort answered all 18 runs correctly, while low answered 15 of 18 using about an eighth of the output tokens.

AWS described three kernel-level optimizations for fused Mixture-of-Experts inference that together delivered up to 10 percent higher throughput and up to 9 percent faster time to first token on H200 GPUs. Two of the changes were upstreamed to vLLM, in versions 0.24.0 and 0.22.0.

AWS now offers OpenAI GPT-6 Sol, GPT-6 Luna and GPT-6 Astra on Amazon Bedrock through global cross-Region inference, letting teams calling from Singapore, Jakarta, Malaysia, Thailand and Taipei route requests to any commercial Region where the model is deployed. Billing, quotas, CloudTrail and CloudWatch stay in the source Region.

AWS has published a walkthrough for serving llama.cpp decision models on Amazon SageMaker AI, including a SemIf adapter that returns probabilities over caller-declared options instead of generated text. A benchmark of 231 JevBench tasks showed a GPU endpoint at 0.352s p50 versus 1.715s on CPU.

AWS has introduced a multi-node Ray LLM Deep Learning Container that bundles vLLM, Ray Serve LLM, NIXL and the EFA stack so one large language model can be split across two GPU nodes on Amazon EKS behind a single OpenAI-compatible endpoint. The walkthrough deploys Qwen3.5-9B with prefill on one node and decode on the other.

AWS's September 2026 recap covers new Bedrock model availability including OpenAI's GPT-6 Astra, Sol and Luna, Anthropic's Claude Fable 5.1 and Opus 5.5, plus general availability of Moonshot AI's Kimi K3 and xAI's Grok 4.6 and 4.7.

Postman's Agent Mode runs on Amazon Bedrock and uses dynamic tool selection that narrows over 170 tools to about 15 per task, plus schema-based reads and tiered prompt caching. Postman reports tool-selection errors increase once the visible toolset exceeds roughly 40 tools.

Incarna used Amazon Bedrock AgentCore payments to let its AI agents pay BlockRun per inference request over x402, completing the integration in three days with roughly 200 lines of code instead of the two to three months originally scoped.

AWS published a reference architecture for sharing one Amazon SageMaker HyperPod EKS cluster across multiple teams using AWS IAM Identity Center, per-team SageMaker AI domains, Kubernetes namespaces, HyperPod Task Governance and namespace-level cost allocation.

Claude Haiku 5.5 is now available on Amazon Bedrock and Claude Platform on AWS, priced around 75 percent lower than Claude Haiku 4.5 for most tasks. Anthropic describes it as the fastest and most efficient model in the Claude 5.5 family.

Amazon Quick and Amazon Bedrock Knowledge Bases now check document permissions with the authoritative source at query time, adding real-time ACL verification on top of existing pre-retrieval filtering. Mondelēz International has deployed Amazon Quick for more than 35,000 employees across four regions.

An AWS Machine Learning Blog post introduces the Agentic Value Model, a framework for building business cases for agentic automation that captures four value dimensions the RPA-era ROI model misses. It also presents a prioritization matrix and cites three Amazon Quick Automate deployments at Kitsa, dLocal, and Genpact.

Qlik built Qlik Answers, a grounded enterprise AI assistant, on Amazon Bedrock for its 40,000+ customers, with general availability in February 2026. Qlik's Discovery Agent has surfaced more than 100,000 discoveries since launch.

AWS published a workflow combining AWS DevOps Agent, Amazon EventBridge, AWS Lambda Durable Functions, and Amazon Bedrock to turn incident investigation summaries into pre-validated remediation actions requiring a single human approval. Read-only actions run autonomously; mutating changes suspend until approved. A demo used a Lambda timeout incident, raising the timeout from 3 to 30 seconds.

An AWS Machine Learning Blog post describes a six-week internal program that paired non-engineering business professionals with mentors and production-grade AWS tools to build working AI agent prototypes. It reports before-and-after self-assessment gains and details one team's WealthWise multi-agent financial advisory prototype.

Cornerstone OnDemand built a multi-agent AI system called Orion AI on Amazon Bedrock and Strands Agents that cut database diagnosis time from 45 minutes to 10 minutes, a 78% reduction. A three-person team delivered the system in six months.

An AWS Machine Learning Blog post describes building a context-aware personal assistant on Amazon Bedrock AgentCore runtime using OpenClaw, an open source agentic system. The example assistant, Sprout, uses AgentCore memory with extraction strategies and metadata filtering, runs from a single CloudFormation template, and routes text and image turns to different Claude models.

AWS published guidance on how customers can align AI governance with the ISO/IEC 42005:2025 standard, focusing on AI system impact assessments. The post describes AWS tools including its ISO/IEC 42001 implementation guide and Responsible AI Lens, and notes AWS holds ISO/IEC 42001 accredited certification for Amazon Bedrock, Amazon Q Business, Amazon Textract, and Amazon Transcribe.