# AWS reports 89.0% success for Continuum on code security benchmark

_Published Monday, October 5, 2026 at 5:54 PM EDT · Security, AI · Latest · Tier 2 — Notable_

AWS says its Continuum security system passed 819 of 920 tasks on CyberGym-E2E within the benchmark's 90-minute limit, achieving an 89.0% success rate. That exceeds the previous public high of 65.9% by 23.1 percentage points.

The benchmark requires agents to find a vulnerability, demonstrate it with an input that crashes the program, and produce a repair that preserves behavior covered by functionality tests. Continuum uses specialized agents for discovery, validation and remediation.

The main success measure does not require fixing the specific historical flaw selected for each task. The benchmark covers memory-safety vulnerabilities in C and C++ projects and excludes deployment context used to prioritize production risks.

## Sources

- [AWS Security Blog](https://aws.amazon.com/blogs/security/aws-continuum-sets-a-new-standard-in-autonomous-code-security/)

---
Canonical: https://techandbusiness.org/newswire/UN9Re4lZ1_kLDCDZ4ZgL1F
Published: 2026-10-05T21:54:52.984Z
Story chronology: 2026-10-05T21:41:01.000Z
Retrieved: 2026-10-05T23:44:02.555Z
Publisher: Tech & Business (techandbusiness.org)