Competitions

Secure Your AI competitions

As a follow-up of the project, we organize a series of challenges covering different AI security aspects (i.e., data leakage, data poisoning, model theft, trojan horse attack, overreliance, and misinformation) and data modalities (satellite telemetry, natural language, and computer vision).

Trojan Horse Hunt in Time Series Forecasting (29.05.2025 - 29.08.2025)

Main task: To identify triggers hidden in AI models for time series forecasting

Space application area: Spacecraft telemetry monitoring

Covered risks from our Security Catalogue: Trojan horse attack, Data leakage, Data poisoning, Model theft

Kaggle link: https://www.kaggle.com/competitions/trojan-horse-hunt-in-space

Prizes: 1000 USD sponsored by KP Labs, ESA merchandise, Guided ESOC tour

Sumarized at: 4th IEEE Conference on Secure and Trustworthy Machine Learning

Summary paper: https://arxiv.org/abs/2603.20108

Stats: 1396 entrants, 217 teams, >40 countries, 1520 submissions

Fake or Real: The Impostor Hunt in Texts (23.06.2025 - 23.09.2025)

Main task: To detect manipulated texts

Space application area: Intelligent assistants and chatbots

Covered risks from our Security Catalogue: Data poisoning, Overreliance

Kaggle link: https://www.kaggle.com/competitions/fake-or-real-the-impostor-hunt

Prizes: 900 USD sponsored by the Centre of Credible AI, ESA merchandise

Summarized at: 4th IEEE Conference on Secure and Trustworthy Machine Learning

Stats: 3295 entrants, 992 teams, >80 countries, 8095 submissions

Space Debris Defence (22.04.2026 - 22.07.2026)

Main task: To fix a model for space debris detection

Space application area: Space debris monitoring, Space situational awareness

Covered risks from our Security Catalogue: Data poisoning, Trojan horse attack

Kaggle link: https://www.kaggle.com/competitions/neural-debris-removal-in-streak-detection-models

Prizes: 1000 USD sponsored by Sybilla Technologies

To be summarized at: EASi-EXPLIMED workshop at IJCAI-ECAI 2026 conference