AI Hacking: Think Like an Attacker to Secure AI
Generative AI and AI agents are transforming how organisations operate, but they are also creating entirely new attack surfaces. Prompt injection, tool abuse, and sensitive data leakage are no longer theoretical risks; they are techniques being exploited in the wild against real-world AI applications.
This hands-on training course takes participants into the mindset of an AI attacker. Drawing on Will's real world experience assessing and exploiting AI systems, you'll learn how modern AI applications can be manipulated and what organisations can do to build resilient AI solutions.
This session is aimed at security professionals, developers, or architects, and you'll leave with an understanding of how attackers target AI systems, the most common AI security vulnerabilities, and recommendations on what your organisation can do to further harden your AI systems.
Participants need to bring a laptop.
Trainer – Will Frame
Will is a penetration tester in the Offensive Security Services team as Bastion. He has a range of experience in pentesting, with a focus on web applications, APIs, and AI/LLM solutions. Will also runs an infostealer monitoring service called restealer (restealer.com). You can find his other stuff at https://wpf.nz
Location:
Bastion
10 Brandon Street
Wellington