AI Jailbreaking: Social Engineering for LLMs - David Willis Owen скачать в хорошем качестве

AI Jailbreaking: Social Engineering for LLMs - David Willis Owen

🎤 Talk Title: AI Jailbreaking: Social Engineering for LLMs 👤 Speaker: David Willis-Owen 📝 Abstract: "AI Jailbreaking" is the process of manipulating a GenAI model to produce unintended outputs. Nearly 3 years after the advent of ChatGPT, Jailbreaks are still feasible and even gaining in popularity. But how are these attacks still possible in 2025? Moreover, how does this impact the Agentic AI future forecasted by so many industry leaders? In this presentation, I will cover the following topics: How & why AI Jailbreaks work Common jailbreaking techniques... & LIVE DEMOS against LLMs! Real-world dangers of Jailbreaks & feasible mitigations The talk is highly topical and very engaging. Don't miss out! ⚓ This talk was recorded live at BSides Bournemouth 2025 on 16th August 2025 — a community-driven cybersecurity conference bringing together researchers, practitioners, and enthusiasts to share knowledge, skills, and ideas. 🌐 Learn more: https://bsides-bournemouth.org/ 💼 Connect with us:   / bsid.  . 📺 Stay tuned for more talks from the event, and don’t forget to subscribe for updates!