CVE-2026-10097
HighCVSS 7.5Exploitation Probability (EPSS)
Low risk5th percentile — higher than 5% of all known CVEs
Summary
The AVX2-optimized ML-KEM implementation in wolfSSL compares only 1536 of the 1568 ciphertext bytes during the Fujisaki-Okamoto re-encryption check in ML-KEM-1024 decapsulation. Ciphertexts differing solely in bytes 1536-1567 bypass implicit rejection and are accepted as valid, breaking IND-CCA2 security.
Risk Assessment
An attacker who can submit chosen ciphertexts to a decapsulation oracle using a static ML-KEM-1024 key and observe whether the genuine shared secret or the implicit-rejection secret is produced can use this as a plaintext-checking oracle to recover the private key. A proof of concept recovered a full ML-KEM-1024 private key with approximately 98% success using roughly 350 chosen ciphertexts.
Recommendation
Immediately update wolfSSL to a version that fixes the bug in the mlkem_cmp_avx2 function to compare all 1568 ciphertext bytes. Until the update, disable AVX2 optimizations for ML-KEM-1024 or use an alternative implementation.
Original NVD description (English source)
wolfSSL's AVX2-optimized ML-KEM implementation (mlkem_cmp_avx2) compares only 1536 of the 1568 ciphertext bytes during the Fujisaki-Okamoto re-encryption check in ML-KEM-1024 decapsulation. Ciphertexts that differ from the expected re-encryption solely in bytes 1536-1567 bypass implicit rejection and are accepted as valid, breaking IND-CCA2 security. An attacker able to submit chosen ciphertexts to a decapsulation oracle that uses a static ML-KEM-1024 key, and to observe whether the genuine shared secret or the implicit-rejection secret was produced, can use this as a plaintext-checking oracle to recover the private key. A proof of concept recovered a full ML-KEM-1024 private key with approximately 98% success using roughly 350 chosen ciphertexts. The flaw is a deterministic logic error and does not rely on timing measurements.