CVE-2026-74874

Use of non-cryptographic random for pixel selection allows data extraction.

• CVSS 8.7
• 338
• Cryptographic
• Remote

openssl_encrypt versions before 1.4.0 use Python's non-cryptographic random module for steganographic pixel selection in the generate_pseudorandom_sequence function. Attackers who know the password can recover the Mersenne Twister state from approximately 624 outputs and predict pixel locations containing hidden data for extraction.

CVE

CVE-2026-74874

CWE

338

CVSS base score

8.7

Published

2026-08-17

OWASP

A02 Cryptographic Failures

Orthogonal defect classification

Algorithm

Code defect classification

Incorrect Algorithm

Category

Cryptographic

Subcategory

Inadequate random number generation

Accessibility scope

Remote

Impact

Information Disclosure

Fixed by upgrading

Yes

Solution

Upgrade openssl_encrypt to version 1.4.0 or later.

import random
from typing import List

def generate_pseudorandom_sequence(password: str, num_pixels: int) -> List[int]:
    """
    Generates a deterministic sequence of pixel indices for steganography.
    """
    # Seed the PRNG with the password to make the sequence deterministic.
    random.seed(password)

    indices = list(range(num_pixels))

    # Use Fisher-Yates shuffle algorithm to generate the sequence.
    for i in range(num_pixels - 1, 0, -1):
        # VULNERABLE: Uses Python's non-cryptographic PRNG which is predictable.
        j = random.randint(0, i)
        indices[i], indices[j] = indices[j], indices[i]

    return indices

import hashlib
import hmac
from typing import List

def generate_pseudorandom_sequence(password: str, num_pixels: int) -> List[int]:
    """
    Generates a deterministic sequence of pixel indices for steganography.
    """
    # Derive a secure key from the password using a KDF.
    key = hashlib.pbkdf2_hmac(
        'sha256', password.encode('utf-8'), b'steganography-salt', 100000
    )

    indices = list(range(num_pixels))

    # Use Fisher-Yates shuffle algorithm to generate the sequence.
    for i in range(num_pixels - 1, 0, -1):
        # FIX: Use an HMAC-based generator for cryptographically secure random numbers.
        h = hmac.new(key, i.to_bytes(4, 'big'), hashlib.sha256)
        j = int.from_bytes(h.digest(), 'big') % (i + 1)
        indices[i], indices[j] = indices[j], indices[i]

    return indices

@misc{vaitp:cve202674874,
  title        = {{Use of non-cryptographic random for pixel selection allows data extraction.}},
  author       = {Bogaerts, Fr\'ed\'eric and Ivaki, Naghmeh and Fonseca, Jos\'e},
  year         = {2026},
  note         = {VAITP Python Vulnerability Dataset, entry CVE-2026-74874},
  howpublished = {\url{https://netpack.pt/vaitp/vulnerability/CVE-2026-74874/}}
}