Understand this story
On November 20, 2026, the MiniMax research team used MiniMax Cyber to conduct automated cryptanalysis of modern cryptographic schemes, uncovering several previously unknown structural flaws and turning multiple theoretical attacks into executable, real-world exploits. The system achieved collision, forgery, or key recovery attacks against former NIST candidates, peer-reviewed designs, and public cryptographic challenges, solving some tasks hundreds to tens of thousands of times faster than the previous state of the art. The research covered only academic designs, former competition candidates, and public challenge parameters, and does not affect deployed cryptographic standards; a technical report and proof-of-concept code have been released.
AI-generated from public reports
Developments
1 confirmed developments
The MiniMax Research Team Uses MiniMax Cyber for Automated Cryptanalysis
Source:MiniMax Blog
Reports and discussion timeline
Browse reports and discussion by time. Their evidence levels remain distinct.
Nov 20
MiniMax BlogFirst-partySelected
MiniMax Cyber Uncovers Multiple Structural Flaws Through Automated Cryptanalysis and Accelerates the Solution of Classic Problems
The MiniMax research team used MiniMax Cyber to analyze modern cryptographic schemes, uncovering multiple previously unknown structural flaws and turning several theoretical attacks into executable, practical exploits. Across historical NIST candidates, peer-reviewed designs, and public cryptographic challenges, it achieved collisions, forgeries, or key recovery, while accelerating some solving tasks by hundreds to tens of thousands of times over the previous SOTA. The research covers only academic designs, historical competition candidates, and public challenge parameters and does not affect deployed cryptographic standards; the technical report and proof-of-concept code have been released at https://github.com/MiniMax-AI/Cyber.
AI Trends