Advanced SOC Operations / CSOC , Events , Next-Generation Technologies & Secure Development
Securing Attack Surfaces With Cyber-Aware Machine Learning
Carnegie Mellon CERT's Clarence Worrell on Role of Machine Learning in Security
As industries embrace digital transformation, machine learning is emerging in many ways across the whole threat detection process, enhancing both speed and accuracy. Clarence Worrell, senior data scientist, CERT Division of Carnegie Mellon University's Software Engineering Institute, highlighted ML's practical applications and emerging challenges in cybersecurity.
See Also: Safeguarding Election Integrity in the Digital Age
Worrell underscored the technology's potential to automate processes and improve security measures within organizations. This automation is evident in diverse applications, from detecting compromised accounts to augmenting the capabilities of SOC analysts through AI-driven tools.
"For machine learning, we are well past the hype cycle. It's going into production, and businesses are realizing value," he said. "Now for generative AI, on the other hand, and large language models, we are in the hype cycle -in the thick of it."
In this video interview with Information Security Media Group at RSA Conference 2024, Worrell also discussed:
- The challenges of "cyber-informed machine learning";
- The explainability issue in ML, particularly in sensitive domains;
- The link between explainable AI and responsible AI principles.
At CERT, Worrell researches data-driven analysis and modeling of cybersecurity. Prior to CERT, he developed applications of machine learning, optimization and probabilistic modeling for the energy sector.
