Flaw in Gemini CLI AI coding assistant allowed stealthy code execution

Flaw in Gemini CLI AI coding assistant allowed stealthy code execution

A vulnerability in Google’s Gemini CLI allowed attackers to execute malicious commands and exfiltrate data from developers’ computers through allowlisted programs. Google released a fix in version 0.1.14 after security firm Tracebit discovered the flaw. #GoogleGemini #Tracebit #CLIExploits

Keypoints

  • A security flaw in Google’s Gemini CLI enabled silent execution of malicious commands.
  • The vulnerability involved processing of ‘README.md’ and ‘GEMINI.md’ files for prompt injection.
  • Attackers could hide malicious instructions within allow-listed commands, leading to data exfiltration.
  • Google issued a security update in version 0.1.14 to address the issue.
  • Users are advised to avoid untrusted codebases and run Gemini in sandboxed environments.

Read More: https://www.bleepingcomputer.com/news/security/flaw-in-gemini-cli-ai-coding-assistant-allowed-stealthy-code-execution/