(S01) (THREAT) AI Use in Cyber Security: A Skeptical Fascination

How do we separate genuine utility from hype, and how do we prepare for the next wave of
adversarial AI in the field. It feels like every vendor has some story of how AI is now embedded
in security products. The criminals have stories of how they are using AI to advance their attacks
too. Who is wining? AI bots vs. AI anomaly detection capabilities, AI stealthy exploitation vs.
AI assisted threat hunting, LLM enabled attacks vs. automated incident response… Beneath
marketing promises lies a race for advantage. Is AI allowing the defenders to close the gap or is
it making the criminals more efficient. There continues to be questions about accuracy, bias,
explainability, and resilience against adversarial manipulation. This presentation takes a critical
look at the first wave of AI and LLM enabled attacks. We’ll take a look at where AI genuinely
strengthens defensive capabilities and where it introduces new attack surfaces, operational blind
spots, or dependence on opaque models. The biggest question of all… If AI is giving the
criminals an even bigger advantage how do we gain control and restrict access?