CTU is taking on nine top universities in Amazon’s latest global AI race, building a system that generates secure, attack-resistant code.

We’re thrilled to share some exciting news – Czech Technical University has been selected as one of only ten universities worldwide to participate in the Amazon Nova AI Challenge! This prestigious competition brings together top academic teams to tackle one of the most pressing challenges in AI today: how to make AI-assisted software development safe and reliable.

"The participation of CTU Prague in this challenge not only showcases the caliber of AI talent in the Czech Republic but also contributes to our goal of fostering innovation in generative AI. We look forward to seeing the innovative solutions that will emerge from this collaboration," said Jiri Sverak, Public Sector Manager at Amazon Web Services in Czech Republic.

AI is revolutionizing the way we write code, but with great power comes great responsibility. AI-generated code needs to be secure, trustworthy, and resistant to attacks – and that’s exactly what this competition is all about. Amazon has designed the Nova AI Challenge as a dynamic tournament where teams either develop AI models to generate secure code (defensive teams) or try to break those models by finding vulnerabilities (offensive teams). It’s AI vs. AI in the best way possible!

Meet Our Team: AlquistCoder

Our team, AlquistCoder, brings together brilliant minds from the Czech Institute of Informatics, Robotics, and Cybernetics and the Faculty of Electrical Engineering. Together, we’re developing an AI system that generates secure programming code (such as Python) while defending itself against adversarial attacks. 

“Our strategy? It’s top secret for now, but let’s just say it involves cybersecurity expertise, extensive testing, and a few tricks up our sleeves,” teases Sebastian Garcia, the team's Faculty Advisor .

Team Members
- Ondrej Kobza (Team Leader)
- Adam Černý
- Maria Rigaki
- Muris Sladić
- Ivan Dostál
- Jan Šedivý (Faculty Advisor)
- Sebastian Garcia (Faculty Advisor)

This isn’t our university’s first rodeo with Amazon competitions. Since 2017, the CTU’s former team (Alquist AI) has competed in multiple AI challenges, including the Alexa Prize Socialbot Grand Challenge, where they took home the gold in 2021. At AIC, we’re happy to expand their original team with our cybersecurity know-how.

AlquistAI's documented success in the 2021 Alexa Prize Socialbot Grand Challenge where they won gold.

The Competition Format

Amazon has set up an intense AI battleground: five universities (including us!) are building AI models, while five others are trying to break them. Every defensive team faces off against each offensive team, making for an exhilarating, high-stakes challenge. The first round kicked off in January 2025, with the final showdown set for June 2025.

And what’s at stake? Besides bragging rights and advancing AI security research, each team receives a $250,000 grant, AWS credits, and the chance to win an additional $250,000. But beyond the financial incentives, the real reward is pushing the boundaries of AI safety and contributing to the future of secure software development.

Defensive Teams (AI Model Development): Carnegie Mellon University, Columbia University, Czech Technical University in Prague, University of Illinois Urbana-Champaign, Virginia Tech

Offensive Teams (Security Testing): NOVA School of Science and Technology in Lisbon, Purdue University, University of California in Davis, University of Texas in Dallas, University of Wisconsin-Madison

Follow Our Journey!

We’ll be sharing updates, insights, and behind-the-scenes stories as we compete in this thrilling challenge. Stay tuned for more, and wish us luck as we take on some of the world’s best AI teams!