An Van Leuven
An is a cybersecurity consultant based in Belgium who has never quite managed to stay on one side of the fence. Her work spans incident response, forensics, threat modelling, cyberdefense projects, and penetration testing, which means she spend roughly equal time thinking about how attacks happen and how organisations survive them.
She spoke at Annacon in 2025 and became genuinely obsessed with what AI actually does to that balance, from both ends of the kill chain.
Session
I built a homemade offensive pipeline from free tools, a few hundred lines of Python, and an LLM that reads attack data and tells you where to strike next. In this talk I show what it can do, where it fails, and what that gap reveals about the future of offensive security. Every tool is free, every concept is reproducible, and the answer to the title question is more interesting than you expect.