Threat · curated 17 Aug 2026
Red Agent Exploits Snowflake Vuln Created by Copilot Autofix
First reported · updated · 2 reports wiz.io
Coverage timeline
Why it matters
The Snowflake incident demonstrates a full AI-versus-AI loop in the software supply chain—an AI coding assistant introducing an exploitable vulnerability and an autonomous offensive AI agent finding and weaponizing it without human intervention—signaling that human code review alone is insufficient as AI-generated code proliferates.
Wiz's autonomous "Red Agent" AI attacker discovered and exploited a GitHub Actions script injection vulnerability in Snowflake's snowflake-connector-net repository, exfiltrating Jira credentials via a crafted issue title. The flaw was inadvertently introduced five days earlier by GitHub Copilot Autofix, which removed a sanitized input pattern (PR #1218) and replaced it with direct string interpolation in a run: block. Snowflake patched it the same day it was reported through its HackerOne program and rotated the affected credential.