In September 2023, the cybersecurity analytics company Splunk, which has never been profitable in its 20 years, was acquired by Cisco Systems for $28 billion. This highlights the gradual merging of two technology ecosystems: information security and data protection with artificial intelligence and machine learning. AI plays an increasingly crucial role benefiting cybersecurity, especially with regards to helping to scale security efforts, threat detection, prediction of advanced persistent threats and ransomware threats, phishing detection, and other domains. However, bad actors, increasingly partnered with nation states, also can access AI and machine learning tools to break defense and wreak havoc. "Jailbreak" versions of Large Language Models or LLM's can be used to rapidly generate code against databases of known vulnerabilities, rendering the possibilities of zero-day exploits more of a threat than ever. Moreover, AI models can generate false positives (such as flagging benign activities as malicious) and false negatives (for instance, missing actual threats). Over-reliance on AI without human oversight can lead to mistakes. It's essential to maintain a balance between automation and human judgment. While AI is a vital tool for modern cybersecurity, providing advanced capabilities and efficiencies, like any tool, it must be analyzed with care, used judiciously and in conjunction with other methods and technologies.

Owen McNally is a professor, researcher, and technology analyst in Austin, Texas, a Director with the National CyberWatch Center, and the EFF-Austin cybersecurity working group organizer. An enduring theme in his work is an evaluation of technologies for usability, quality, and ethics based on humanistic principles. He holds an interdisciplinary Ph.D. from the University of Texas at Austin, with a program of work in Medical Cognitive Science. He has taught college-level software design, cognitive studies, psychology, research methods, and statistics for the behavioral sciences. In recent years his work has been in cybersecurity readiness, hospital software, modeling of COVID dynamics in populations, software design and usability, risk analysis, and researching investment opportunities in AI, neurotechnology, logistics and other sectors.

Talk will both be livestreamed here and later archived here on the EFF-Austin YouTube channels. Questions for the speaker from virtual attendees may be submitted via the Youtube livestream chat, our Twitter account, @EFFaustin, or our Mastodon account, @effaustin@mastodon.social.