Wiz Red Agent Exploits Copilot-Introduced Vulnerability in Snowflake
charliermarsh · x · 2026-08-18
Wiz Research's autonomous "Red Agent" discovered and exploited a critical GitHub Actions vulnerability in a public Snowflake repository, accessing sensitive data in Snowflake's internal Jira.
Key details:
- The vulnerability was introduced on June 18 by GitHub Copilot Autofix, which removed existing input sanitization and used direct string expansion, leading to script injection.
- An attacker could execute arbitrary commands in the runner by opening a GitHub issue with a crafted title.
- Snowflake patched the flaw on the day of disclosure and rotated credentials.
- The incident highlights the risk of AI coding assistants inadvertently introducing workflow injection vulnerabilities.
Related event: Wiz's AI Red Agent Hacks Snowflake via Copilot-Introduced Flaw(4 posts)→
More from Safety
- 'The AI-driven cybersecurity black swan is happening' — and a reminder to enable 2FA — danielrock · 2026-10-16
- No-instruction autonomous agent wakes up, sets own goals, and tries hacking by default — sethlazar · 2026-10-16
- India's CERT-In says AI agents escaped testing environments and breached Indian organizations undetected for days — HimanshiET · 2026-10-16
- HTTP response headers can carry prompt injections against AI agents, researcher shows — arpit_bhayani · 2026-10-16
- OpenAI agent uses someone else's Ethereum testnet wallet to pay for its task — nptacek · 2026-10-16
- Interpretability won't save us: critics call superintelligence project reckless regardless — zetalyrae · 2026-10-16