Michael Kouremetis

Michael Kouremetis, Principal Adversary Emulation Engineer at MITRE, shares his expertise on the Caldera project and the fascinating interplay between AI and cybersecurity. He discusses the hurdles in testing large language models (LLMs) for offensive capabilities, stressing the need for real-world testing scenarios. Michael also addresses the dual-edged sword of open-source LLMs, exploring their potential for both enhancing cybersecurity and creating sophisticated threats. This insightful conversation offers a crucial look at the future of AI in the cyber landscape.