CVE-2024-12366

PandasAI uses an interactive prompt function that is vulnerable to prompt injection and run arbitrary Python code that can lead to Remote Code Execution RCE instead of the intended explanation of the natural language processing by the LLM...