Hacking-Back the AI-Hacker - Defending Against LLM-Cyberattacks the Cool Way

Learn how to defend against AI-driven cyberattacks through an innovative counter-hacking approach in this 47-minute conference talk from RSA Conference. Discover how large language models are making cyberattacks more accessible, scalable, and cost-effective for malicious actors. Explore Mantis, a groundbreaking defensive system that serves as a first line of defense against AI-powered cyberthreats by exploiting inherent vulnerabilities within attacking LLMs themselves. Understand how this defensive mechanism works by injecting strategically crafted prompts into system responses that cause the attacker's LLM to sabotage itself or even engage in self-hacking behaviors. Gain insights into this cutting-edge cybersecurity strategy that turns the attackers' own AI tools against them, presented by experts from the field including a Principal Researcher and CEO specializing in AI security solutions.