At the AWS Summit held in New York, Amazon Web Services (AWS) announced the launch of two new services—AWS Continuum and AWS Context—designed to address two major bottlenecks in deploying AI agents in production environments: security risks and lack of business context.

Amazon, AWS

AWS Continuum focuses on automating the repair of code vulnerabilities, prioritizing business impact for both publicly known and newly scanned risks, and automatically executing remediation measures after validation in an isolated environment. AWS Context, on the other hand, builds a shared knowledge graph using enterprise data, providing all agents within the organization with context-aware capabilities to prevent intelligent agents from fabricating incorrect information.

In addition to the new services, AWS has also upgraded its existing ecosystem. In response to the background of service disruptions caused by AI coding tools in February 2026, which led to stricter code review policies internally, the AWS DevOps Agent in the testing phase now includes a release readiness review and dynamic test plan generation feature. This preview version is currently available free of charge in the US East region.

QQ20260622-101135.jpg

Meanwhile, the coding agent Kiro launched a native iOS app to enable cross-platform synchronization control; the Bedrock AgentCore, a platform for managing production agents, has integrated mainstream third-party data connectors and security filters. A series of密集 product launches and functional updates by AWS reflect the clear trend that AI agent technology is accelerating from "experimental testing" to "production-level applications." Building a comprehensive security defense and deep business context connections has become a key foundation for the successful implementation of enterprise-level AI agents.