AI + Security
Building defensive AI systems and securing the future of machine learning.
Overview
As AI becomes integrated into every layer of our digital infrastructure, the need for robust security is paramount. This track focuses on the intersection of AI and Cybersecurity—both using AI to defend and securing AI itself from adversarial attacks.
Defensive AI
Systems that detect and respond to threats in real-time.
Model Robustness
Securing LLMs and ML models against prompt injection and data poisoning.
Automation
Using AI to automate tedious security auditing and compliance tasks.
What we're looking for:
- Practical tools that solve real security problems using AI.
- Novel research into LLM security and jailbreaking prevention.
- Efficient monitoring systems for large-scale infrastructure.
- Seamless integration into existing developer workflows.
Judging Criteria
Technical Sophistication
How complex and well-engineered is the AI implementation? Does it move beyond simple API wrappers?
Security Impact
Does the project significantly improve the security posture of a system or model?
Novelty
How unique is the approach? Does it solve a problem in a way that hasn't been done before?
Usability
Can a security professional or developer actually use this tool in their daily work?