The Tech TL;DR:

• Agentic CX Designer: A no-code canvas for Amazon Connect Customer that lets business teams deploy AI-powered voice and chat experiences.
• AWS Lambda MicroVMs: Firecracker-backed virtual machine isolation designed for running untrusted user or AI-generated code with near-instant resume times.
• EC2 AMI Watermarks: Native tagging for private Amazon Machine Images that automatically persists across copies, regions, and account shares.

Serverless Compute Primitives and MicroVM Isolation

Per the official AWS release notes, AWS Lambda MicroVMs arrive as a new serverless compute primitive built on Firecracker virtualization technology. The architecture provides individual virtual machine isolation for each user or job, supporting near-instant launch and resume speeds alongside the capability to suspend and resume execution states for up to eight hours. For distributed systems engineers managing multi-tenant applications executing untrusted AI-generated code, this abstraction removes the need to provision underlying hypervisor infrastructure while maintaining strict workload boundaries.

# Example payload configuration for a multi-tenant Lambda function
{
  "FunctionName": "tenant-executor-core",
  "Runtime": "provided.al2023",
  "MemorySize": 1024,
  "EphemeralStorage": {
    "Size": 512
  },
  "Timeout": 30,
  "Architectures": ["arm64"]
}

Governance, Security, and Image Compliance

To address compliance auditing across multi-account AWS environments, Amazon EC2 AMI Watermarks now enable developers to embed custom identifiers directly into private AMIs. According to AWS documentation, these watermarks automatically propagate to every derived machine image across cross-account shares and regional copies. Security teams can pair these watermarks with Allowed AMIs and Declarative Policies to enforce strict pipeline validation.

On the observability front, Amazon GuardDuty launched AI-powered investigations in preview. The system analyzes raw telemetry and historical account findings across a 90-day window utilizing knowledge graphs and threat intelligence to isolate genuine security threats from benign activity. Each generated investigation surfaces a disposition assessment complete with confidence scoring and MITRE ATT&CK classification mappings.

Ecosystem Migrations and Open Governance Shifts

Database modernization efforts received a boost with Amazon OpenSearch Service introducing AI-assisted migrations. The Migration Assistant now features an agent-guided workflow built to streamline transitions from self-managed Apache Solr, Elasticsearch, or OpenSearch deployments to managed or serverless architectures. This rollout includes live traffic capture and replay capabilities specifically engineered for Solr workloads. Similarly, Amazon MSK AI Agent Skills now provide specialized guidance for coding assistants like Kiro, Claude Code, and Cursor to handle cluster sizing, migration, and troubleshooting directly within developer IDEs.

In broader open-source governance news, Oracle announced a community governance model for MySQL, establishing a Steering Committee that reserves four non-Oracle seats. AWS confirmed it holds one of these seats, continuing its upstream contributions to the database project. Meanwhile, AWS announced that administrators can maintain eligible certifications for an additional year through curated labs on AWS Skill Builder rather than re-taking full proctored exams.