Description

Detects insecure deserialization vulnerabilities when using Python's pickle.loads() function with untrusted input. The pickle module is inherently unsafe for deserializing untrusted data as it can execute arbitrary code during deserialization, potentially leading to remote code execution attacks.

Detection Strategy

• Check if the 'pickle' module is imported in the Python source code

• Look for calls to functions ending with 'loads' (targeting pickle.loads)

• Examine if the first argument to loads() contains or is derived from user-controlled input

• Verify the input argument is not properly sanitized before being passed to loads()

• Report a vulnerability if unsafe user input reaches pickle.loads() without proper validation

Vulnerable code example

import pickle
from flask import Flask, request

app = Flask(__name__)

@app.post("/api/data")
def process_data():
    raw_data = request.get_data()...

✅ Secure code example

import pickle
from flask import Flask, request, jsonify

app = Flask(__name__)

# Define whitelist of allowed pickle data
SAFE_PICKLES = {
    pickle.dumps({}),...