"Every recommendation includes the reasoning behind it. As you gain confidence, you can graduate Continuum to enforce mode, enabling remediation that can be increasingly automated based on categories and risk profiles you define,” AWS said.
What AWS is offering with Continuum
AWS on June 17 introduced Continuum, a new Amazon-made platform designed to manage the full lifecycle of code vulnerabilities on AWS infrastructure. Announced at AWS Summit New York, Continuum is available in a gated preview and is presented as an integrated system with access to an organization’s full environment — both structured data that lives in AWS and unstructured sources such as documents, communications and business priorities.
Four explicit capabilities: discovery through remediation
AWS describes Continuum as a single platform that follows vulnerabilities from initial discovery to remediation. The company lists four capabilities in sequence: (1) code vulnerability discovery, where Continuum ingests an existing backlog of vulnerabilities and performs its own scans of the environment; (2) code vulnerability prioritization, where Continuum evaluates, enriches and prioritizes findings using contextual information and produces an evidence-backed list of priorities; (3) code vulnerability validation, where Continuum attempts to surface false positives, provides additional context and constructs working exploit examples in a sandboxed environment; and (4) code vulnerability mitigation and remediation, where Continuum assesses existing defenses — including blocking and compensating controls and detection mechanisms — and then recommends mitigation or remediation via network changes, policy changes or code patches.
AWS Security Agent, frontier AI models, and STRIDE output
The Continuum platform includes an AWS Security Agent powered by what AWS describes as frontier AI models. That agent is intended to assist software developers and security engineers with penetration testing, code scanning and threat modelling. AWS said the output from these tools will be presented in the Spoofing, Tampering, Repudiation, Information disclosure, Denial of service, Elevation of privilege (STRIDE) format. AWS has rebranded those capabilities inside Continuum as Continuum pen testing, Continuum code scanning and Continuum threat modelling.
Human-in-the-loop, enforcement mode, and the stated motivation
AWS emphasized that Continuum always begins “in learn mode” with a human in the loop. The company framed that design as deliberate: recommendations include the reasoning behind them, and organizations can choose to “graduate” Continuum to an enforce mode to enable progressively automated remediation according to categories and risk profiles they define. AWS explained the launch as a response to an “urgent need for a shift” in security workflows — shifting from traditional telemetry-and-dashboard models toward “telemetry, context, reasoning and actions.” AWS also linked the urgency to advances in AI, saying that “Models like Claude Mythos can now find software vulnerabilities and reason through complex attack paths at machine-speed, leading to an exponentially increasing backlog of vulnerabilities.”
How security teams, software developers, and enterprise customers are positioned
- Security teams and software developers: Continuum is positioned to change daily workflows by combining backlog ingestion, automated scanning, contextual prioritization and sandboxed validation. The inclusion of the AWS Security Agent and STRIDE-formatted outputs targets both offensive testing (pen testing) and defensive coding and threat modelling tasks for engineers and security practitioners.
- Enterprise customers in financial services, automotive and technology: AWS confirmed customers across these sectors were already using Continuum in its preview. For these organizations, Continuum’s ability to draw on structured and unstructured internal data and recommend network, policy or code-level fixes is presented as a way to reduce a growing remediation backlog.
Continuum is one element among a set of announcements that also included new AI models and an AWS Context knowledge graph intended to give agents access to contextual information. The product is being billed by AWS as a workflow shift — and it is being released into a reality AWS itself describes as one where automated models can rapidly expand the universe of identified vulnerabilities.
Continuum’s preview and the company’s emphasis on human oversight leave concrete questions for adopters: how organizations will calibrate the transition from learn mode to enforce mode, how sandboxed exploit construction will be governed, and how evidence-backed prioritization will map to existing change and risk processes. For now, AWS has placed Continuum in controlled preview while inviting customers across multiple sectors to try a platform that stitches telemetry, context and reasoning into a single product offering.
Original story: https://www.infosecurity-magazine.com/news/aws-continuum-ai-vulnerability/
