Uploaded September 2025 | Updated September 2026, 2 weeks ago
In this talk, Team Atlanta will share their key experiences and insights from participating in DARPA's AIxCC competition. They'll highlight their technical approach to leveraging AI systems—particularly large language models—to identify and remediate software vulnerabilities. The goal is to distill what worked, what didn’t, and what surprised us, offering practical takeaways for researchers and practitioners working with LLMs in security contexts. By reflecting on the challenges and opportunities we encountered, they hope to contribute to the broader conversation on developing AI-first tools to help secure critical infrastructure in the years ahead.
Speaker:
Taesoo Kim, VP at Samsung Research and Professor at Georgia Tech
In this talk, Team Atlanta will share their key experiences and insights from participating in DARPA's AIxCC competition. They'll highlight their technical approach to leveraging AI systems—particularly large language models—to identify and remediate software vulnerabilities. The goal is to distill what worked, what didn’t, and what surprised us, offering practical takeaways for researchers and practitioners working with LLMs in security contexts. By reflecting on the challenges and opportunities we encountered, they hope to contribute to the broader conversation on developing AI-first tools to help secure critical infrastructure in the years ahead.
Speaker:
Taesoo Kim, VP at Samsung Research and Professor at Georgia Tech
