In today’s digital age, critical infrastructure relies heavily on software systems to function efficiently and effectively. From transportation networks to healthcare facilities, keeping this software protected and working flawlessly is vital to the safety and well-being of society as a whole. Cyber-Reasoning-Systems (CRS) have emerged as a powerful tool in the fight against software vulnerabilities, offering automated solutions for patching and securing critical software systems.
Our Team
Our team consists mainly of graduate and postgraduate students at the University of Bonn (Behavioral Security Research Group) and Fraunhofer FKIE, supported by Code Intelligence LLC. We have diverse expertise in software security, machine learning, and automated program analysis, which we combined to build our cyber reasoning system.
AIxCC Semi-Finalists
We designed and implemented a fully automated cyber reasoning system (CRS) that finds, analyzes, and patches real-world software vulnerabilities. In the AIxCC, we reached the semi-finals by leveraging AI to analyze bugs and generate patches, building a robust infrastructure to fully automate the process.
Research & Impact
Our work on large language model based vulnerability detection and patching systems yielded exciting results during the competition. We have since published our findings and presented our learnings from the AIxCC in papers and talks at the 39th Chaos Communication Congress (39C3: Power Cycles).
We participated in the Artificial Intelligence Cyber Challenge (AIxCC), reaching the semi-finals and successfully validating our system. The competition served as a platform for innovation and excellence, facilitating collaboration with top AI companies. We have published our findings and presented our experiences in papers and talks at the 39C3 (2025).
Join us on our mission to secure the software that powers the world’s most essential services. Together, we will ensure that critical infrastructure remains resilient, reliable, and secure for generations to come.