Secret Loyalties Hackathon | Electric Sheep X Apart Research
Join the Secret Loyalties Hackathon organized by Apart Research, co-organised by Formation Research, in partnership with Electric Sheep
As frontier AI systems become increasingly capable, understanding and detecting hidden objectives is becoming one of AI safety's most important research challenges.
Over one weekend, you'll collaborate with researchers and builders to develop model organisms, detection methods, evaluations, and prototype defenses that advance research on secret loyalties in AI systems.
After the Luma registration, please follow the below instructions:
👩🏻💻 Sign up here and review the tracks: https://apartresearch.com/sprints/secret-loyalties-hackathon-2026-07-24-to-2026-07-26
🔗 Join the Discord Channel: Alignment Jams
👭 Participation: This is a fully VIRTUAL event, open to participation across India
🏆 Prize: $2,000 USD prize pool and many more.. *Invitation to the Apart Fellowship for top team, publication pathway and mentor introductions
👥 WHO SHOULD JOIN?
• AI Safety Researchers • Machine Learning Researchers & Engineers • Cybersecurity Researchers • Interpretability & Alignment Researchers • Students • Research-minded Builders
🔬 WHAT YOU'LL BUILD
• Model Organisms • Detection Methods • Evaluations • Threat Analysis • Defenses
No prior AI safety research experience is required, just bring a team and an idea.