AgentsAWS Agentic AI Security Scoping Matrix — Four-Scope FrameworkAWS Security Blog·high signalFour scopes for agent security: read-only, propose changes, autonomous within bounds, self-initiating. Each maps to specific security controls.SourceSource pageAWS Security Blog↳ Follow the threadStack layer / Threat patternAWS documents two WAF architecture patterns for securing Bedrock AgentCore RuntimeAWS Machine Learning BlogStack layer / Threat patternAWS open-sources Loom, an enterprise agent platform built on Strands Agents + Bedrock AgentCore RuntimeAWS Open Source BlogStack layer / Threat patternAWS documents four deployment patterns for Unsloth-quantized models on SageMakerAWS Machine Learning BlogStack layer / Threat patternAWS Adds Serverless Fine-Tuning for NVIDIA Nemotron 3 Models on SageMaker AIAWS Machine Learning BlogStack layer / Threat patternSageMaker HyperPod inference adds multi-tier data capture, direct Hugging Face Hub deployment, and NVMe cachingAWS Machine Learning BlogPolicy dependency / Stack layerIAPS warns on Kimi Claw: security and governance risks of Chinese-hosted 'always-on' AI agentsIAPSThreat pattern / Update threadAWS shows how to build a semantic layer for agentic AI with Stardog on Bedrock AgentCoreAWS Machine Learning BlogPolicy dependency / Stack layerKastra Launches Policy Enforcement for Claude Code, Cursor, and CodexHacker News
AWS Security Bloghigh signal