AWS embeds security into OpenAI, Anthropic coding tools with new Continuum integrations

Amazon Web Services just made a bet that the best way to secure AI-generated code is to catch the problems before they leave the IDE. The cloud giant announced integrations between its AWS Continuum security service and three major AI coding tools: Anthropic’s Claude Code, OpenAI’s Codex, and AWS’s own Kiro.

The announcement, timed to Black Hat USA 2026, essentially turns the security layer into a ride-along for every line of AI-assisted code a developer writes.

What AWS Continuum actually does

AWS Continuum first entered gated preview on June 17, 2026, operating as an AI agent-driven service designed to manage code vulnerabilities within customer-defined guardrails. The new integrations extend that capability directly into the coding tools that developers are already using day to day.

Rather than bolting security onto the end of the development pipeline, Continuum embeds real-time vulnerability detection and remediation recommendations into the workflow itself. When Claude Code or Codex generates a block of code, the security layer evaluates it on the spot and flags issues before they compound downstream.

The practical effect is a much tighter feedback loop within the Development, Security, and Operations framework, commonly known as DevSecOps.

The enterprise problem this solves

AWS’s approach treats this as an infrastructure problem rather than a behavioral one. The guardrails are set by the customer’s security team, meaning organizations can enforce their own policies without relying on individual developer discipline.

The select design partners currently testing the integrations in preview mode include Capital One, MongoDB, Rivian, and Robinhood. That’s a notable roster, spanning financial services, automotive technology, database infrastructure, and fintech.

Strategic positioning in the AI tooling wars

The decision to integrate with both Anthropic and OpenAI coding tools is worth pausing on. AWS has its own AI coding assistant in Kiro, and Amazon has invested heavily in Anthropic as a company. Extending Continuum to OpenAI’s Codex as well signals that AWS is playing the platform game rather than trying to lock developers into a single ecosystem.

The timing also matters. Black Hat USA is the premier cybersecurity conference, and announcing there sends a clear message about AWS’s priorities. This isn’t being positioned as an AI feature. It’s being positioned as a security feature that happens to work with AI tools.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.