AI Security Competition @ ODU

The School of Cybersecurity at ODU invites you to participate in its first AI Security Competition. This Large Language Model (LLM) Capture The Flag (CTF) competition is an exciting event where participants challenge LLMs and test their skills in adversarial prompting. The competition features 10 levels, each hiding a secret flag within different LLMs. Your mission? Use prompts to extract these hidden flags!

🚀 No technical skills required! This competition is open to everyone at ODU (students, faculty, and staff), regardless of major or technical background. All you need is curiosity and a willingness to explore AI security in a fun and hands-on way!

📅 Competition Dates

March 3, 10 AM – March 16, midnight

🏆 Prizes

🥇 First Place: $500

🥈 Second Place: $300

🥉 Third Place: $200

Prizes will be awarded only to students.

🕵️‍♂️ Competition Details

• Each level presents a unique LLM with different security safeguards.
• Participants must craft ingenious prompts to bypass restrictions and reveal the hidden flags. All techniques must be ethical and focused on responsible AI security research.

⚠️ Important

Trying to hack or manipulate systems beyond the allowed challenges is strictly prohibited.

📊 Scoring & Ranking

• Each successfully revealed flag earns a score based on the difficulty level.
• Participants are ranked based on their total score.
• In case of a tie, the participant who used fewer prompts will be ranked higher.

🎓 Who Can Participate?

This competition is open to everyone at ODU (students, faculty, and staff); however, prizes will only be awarded to ODU students. Whether you’re a beginner or an expert, this is your chance to prove your ability in outsmarting AI systems!

To participate, please register using your ODU email.

Do you have what it takes to jailbreak the models and capture all 10 flags?