AWS Unveils Continuum to Detect and Remediate Code Vulnerabilities at Scale

Amazon Web Services (AWS) has announced the launch of AWS Continuum for code vulnerabilities, now available in gated preview.

The platform represents a fundamental shift in enterprise security operations, moving away from passive telemetry dashboards toward an AI-driven system that discovers, prioritizes, validates, and remediates vulnerabilities at machine speed.

Unveiled on June 17, 2026, by Chet Kapoor, AWS Continuum is a model-agnostic, agentic security platform built to handle the full lifecycle of code vulnerabilities.

AWS Unveils Continuum to Detect and Remediate

It ingests both structured data, such as infrastructure configurations, permissions, and network topology, and unstructured organizational data, including documents and business priorities, to reason through threats with full environmental context.

The announcement comes amid growing pressure from next-generation models at the cybersecurity frontier.

Models like Claude Mythos can now autonomously identify software vulnerabilities and map complex attack paths, generating an exponentially increasing backlog that traditional security operations centers simply cannot process manually.

The Discovery phase ingests existing vulnerability backlogs and performs its own environment scans to build a comprehensive attack-path map covering both first- and third-party code.

From there, the Prioritization phase uses business context, deployment status, network reachability, production exposure, and potential blast radius to generate an evidence-backed priority list that helps teams focus on what matters most.

During Validation, Continuum constructs working exploit proofs-of-concept in sandboxed environments to confirm real vulnerabilities and surface false positives before they consume analyst time.

Finally, the Mitigation and Remediation phase recommends and validates network changes, policy updates, or code patches using the same system that confirmed the original vulnerability, with rollback paths included where feasible.

Continuum is designed with a graduated trust model. It initially operates in learn mode, keeping a human in the loop with full reasoning transparency for every recommendation.

As security teams build confidence, they can promote it to enforce mode, enabling increasingly automated remediation scoped to risk profiles and vulnerability categories they define.

This approach ensures that automation expands proportionally with organizational confidence, a critical design choice for enterprise environments. Beyond core vulnerability management, AWS is integrating several existing tools under the Continuum umbrella. 

Continuum Pen Testing incorporates the formerly standalone AWS Security Agent penetration testing module, while Continuum Code Scanning (Preview) delivers automated static analysis from the same platform.

Chet Kapoor stated that AWS is also launching Continuum Threat Modeling (Preview), which automatically generates STRIDE-format threat models directly from design documents or source code.

All three capabilities feed detection signals back into the broader Continuum loop, enriching discovery and prioritization across the board.

AWS Continuum signals a broader industry pivot: security tooling is moving from reactive monitoring toward agentic, outcome-driven platforms.

By embedding business context alongside technical telemetry, AWS is positioning Continuum to reduce mean time to remediate at a scale previously unattainable with human-led workflows alone.

Follow us on Google News , LinkedIn and X to Get More Instant Updates. Set Cyberpress as a Preferred Source in Google.