This talk highlights the disruptive impact of AI in cybersecurity. We'll explore how large language models are utilized by both defenders and attackers in the cyber world, leading to a complex cat-and-mouse game. We will start with some noteworthy threats that were built on AI before we move to the defender perspective. In addition to Microsoft's Security Copilot, we will also talk about general requirements for dealing with large language models before moving on to the question of enablement. In addition to getting started with AI in cybersecurity, there will also be a few best practices that will be your key takeaways.