Google Fixes Critical RCE Flaw in AI-Based Antigravity Tool

Google patched a critical prompt-injection vulnerability in its agentic IDE Antigravity that allowed attackers to inject command-line flags into the underlying fd utility via the find_by_name tool, enabling sandbox escape and remote code execution. Researchers at Pillar Security developed a proof-of-concept showing a full attack chain—staging a malicious script and triggering it through a seemingly benign search—and noted that the vulnerability bypassed Antigravity's Secure Mode because the tool executes before security restrictions are applied; Google fixed the issue after disclosure.