Machine Learning is Best Suited for Security

Although artificial intelligence, specifically machine learning and large language models, is in the news, it isn’t very useful in enterprise IT. In this episode of the On-Premise IT podcast, Karen Lopez, W. Curtis Preston, Michael Levan, and Stephen Foskett discuss the use case for AI in security. The panel acknowledges that machine learning can be beneficial in identifying anomalies and patterns that humans may overlook. It can assist in generating policies, templates, and rule sets, as well as providing best practices based on aggregated data. However, they also express concerns about the responsible use of AI and the need for training models on specific environments to ensure effectiveness. They highlight the importance of having the right data sets and the challenges of dealing with the black box nature of machine learning. Despite potential exploits and limitations, they agree that AI is currently the best tool available for detecting and addressing security threats, such as data exfiltration and unauthorized access.

© Gestalt IT, LLC for Gestalt IT: Machine Learning is Best Suited for Security