Harness Deploys AI Agents to Close the Six-Hour Exploit Window
Attackers now weaponize frontier AI models to exploit software vulnerabilities in as little as six hours, leaving traditional security teams—who often take 50 days to patch—critically exposed. Harness is attempting to bridge this gap by launching a suite of autonomous agents designed to automate the entire vulnerability lifecycle.
The security infrastructure unveiled by the San Francisco-based company integrates scanning, triage, and remediation directly into the software delivery pipeline. By embedding these capabilities, Harness aims to remove the manual handoffs that typically stall incident response. The toolkit includes a Zero-Day Agent that monitors for new threats and generates validated fixes within minutes of disclosure, alongside a virtual patching feature that shields production environments before permanent code changes are finalized.
Rahul Sood, General Manager of Application Security, noted that the platform forces security to operate at the same machine speed as the AI models currently driving software development. According to the company, these tools utilize reachability data to filter out noise and prioritize exploitable risks. This development follows Harness's 2025 merger with Traceable and represents a broader industry shift toward embedding governance and automated response directly into the CI/CD process. The new suite, which includes AI SAST and automated Triage and Remediation agents, is available to customers immediately.
Comments (0)
No comments yet. Be the first!