AI attacker broke into Snowflake's internal Jira via a flaw introduced by Copilot Autofix
shirtamari · x · 2026-08-17
A security researcher's autonomous AI attacker made its way into Snowflake's internal Jira and accessed sensitive data — with nothing more than a public GitHub issue carrying a crafted title.
The twist: the vulnerability wasn't human error. It had been introduced 5 days earlier by GitHub's "Copilot Autofix powered by AI" — an AI-powered auto-fix creating a hole that another AI then exploited. A concrete demonstration of agentic attack surfaces like indirect prompt injection in real supply chains.
More from coding & agent
- Dev runs simulated town where each citizen is a Codex agent task — Dimillian · 2026-08-17
- Netlify integrates OpenRouter to access 400+ models — thisiskp_ · 2026-08-17
- DaedalMap Earthquake Data MCP connector provides global earthquake events — modelcontextprotocol · 2026-08-17
- Ultimate GSAP Master MCP generates production-ready 60fps GSAP animations via natural language — modelcontextprotocol · 2026-08-17
- ElevenLabs Launches MCP Integration to Manage Voice Agents in Claude — lukeharries · 2026-08-17
- OpenAI training models to write superhumanly secure code, says Brockman — tekbog · 2026-08-17