Following the official welcome by CISPA-Faculty Dr. Adam Dziedzic and Dr. Franziska Boenisch from the SprintML Lab, the participants launched into an intensive working phase. Within just 24 hours, the teams developed creative and technically sophisticated concepts, which they later presented to a jury. Special thanks also go to our guests: Michał Zawalski from NVIDIA for his keynote on benchmark contamination—the unintended overlap between training and test data—and Prof. Robert Nowak, Deputy Director of Science at the Institute of Computer Science at Warsaw University of Technology, for the warm welcome.

Challenges Inspired by Real Research Problems

As part of the competition, bachelor’s and master’s students from several countries tackled three hands-on challenges from the fields of artificial intelligence and cybersecurity. The tasks focused on understanding how training data influences the behavior of AI models, how information can be reconstructed from training data, and how AI-generated results can be made more interpretable.

One challenge centered on the question of which sensitive information modern models may unintentionally memorize and reveal from their training data. For this task, CISPA researchers developed a dedicated dataset simulating realistic private user information. Participants were given access to a trained model along with individual clues—for example, an image or fragments of information. Their objective was to use so-called model inversion attacks to reconstruct which personal data the model had “retained” from training and how these fragments could be combined into a complete digital profile. Teams were evaluated based on how accurately their reconstructed datasets matched the original training data.

The challenge highlighted a key security and privacy issue in modern AI systems: Even when training data itself is not directly accessible, models may under certain circumstances unintentionally disclose information about the data they were trained on.

Research into these risks—and the development of secure AI systems—is a central focus of research at the CISPA Helmholtz Center for Information Security.

The Winning Teams from Warsaw

In the end, the three teams (Syntax Terror, zer0_day, Advanced Persistent Thinkers) impressed the jury in Warsaw. Participants brought a wide range of experiences and backgrounds to the competition: While some were only at the beginning of their bachelor’s studies, others already had several years of professional experience in IT. The exceptionally high level of expertise among participants once again demonstrated the strength and dynamism of Poland’s machine learningcommunity, with which CISPA shares a particularly close connection.

Next Stop: the CISPA Grand Finale

The event also marked the conclusion of the six regional qualification rounds of the CISPA European Hackathon Championship, which took place across Europe between November 2025 and June 2026. With the Warsaw hackathon, not only did the final regional event come to a successful close, but so did a Europe-wide series that gave numerous young talents the opportunity to showcase their skills, build international connections, and work on real-world challenges in artificial intelligence and cybersecurity.

Attention now turns to the next major milestone: the CISPA Grand Hackathon Finale. There, the top teams from the regional competitions will once again compete against one another.

The Europe-wide Hackathon Championship was made possible through the commitment of Dr. Adam Dziedzic, Dr. Franziska Boenisch, and the entire CISPA project team, who organized and supported the Hackathon Championship across Europe. We would also like to extend our sincere thanks to all organizers, partners, and supporters who helped make this event series possible.

Most importantly, however, we thank all participants for their dedication, creativity, and team spirit. It has been a pleasure to have you with us on this journey—and we look forward to seeing many of you again at the CISPA Grand Finale.

Our Team in Warsaw

About the Hackathon Championship

The CISPA European Cybersecurity & AI Hackathon Championship is a Europe-wide competition organized by the CISPA Helmholtz Center for Information Security from November 2025 to July 2026. In regional qualifiers held in major European university cities, bachelor’s and master’s students compete in teams of up to four people to solve challenges in the fields of AI and cybersecurity within 24 hours. The winners from each city qualify for the grand finale in St. Ingbert, where they compete for cash prizes, trophies, and certificates. By bringing together young talent from across Europe, this championship aims not only to promote innovation and expertise in trustworthy AI and cybersecurity, but also to build a pan-European community dedicated to securing our digital future.