PyCryptodome and pycryptodomex before 3.19.1 allow side-channel leakage for OAEP decryption, exploitable for a Manger attack.

PyCryptodome before 3.6.6 has an integer overflow in the data_len variable in AESNI.c, related to the AESNI_encrypt and AESNI_decrypt functions, leading to the mishandling of messages shorter than 16 bytes.

lib/Crypto/PublicKey/ElGamal.py in PyCrypto through 2.6.1 generates weak ElGamal key parameters, which allows attackers to obtain sensitive information by reading ciphertext data (i.e., it does not have semantic security in face of a ciphertext-only attack). The Decisional Diffie-Hellman (DDH) assumption does not hold for PyCrypto's ElGamal implementation.

Heap-based buffer overflow in the ALGnew function in block_templace.c in Python Cryptography Toolkit (aka pycrypto) allows remote attackers to execute arbitrary code as demonstrated by a crafted iv parameter to cryptmsg.py.

The Crypto.Random.atfork function in PyCrypto before 2.6.1 does not properly reseed the pseudo-random number generator (PRNG) before allowing a child process to access it, which makes it easier for context-dependent attackers to obtain sensitive information by leveraging a race condition in which a child process is created and accesses the PRNG within the same rate-limit period as another process.

PyCrypto before 2.6 does not produce appropriate prime numbers when using an ElGamal scheme to generate a key, which reduces the signature space or public key space and makes it easier for attackers to conduct brute force attacks to obtain the private key.