NVIDIA recently demonstrated its AI security expertise at two major cybersecurity conferences, Black Hat USA and DEF CON 32. These conferences provided a platform for NVIDIA to showcase its latest advancements in AI security and share valuable insights with the cybersecurity community.

NVIDIA at Black Hat USA 2024

Black Hat USA is a renowned event for cutting-edge security research. At this year’s conference, NVIDIA highlighted the applications of generative AI tools in security and the security of AI deployments. The keynote by Bartley Richardson, NVIDIA’s Director of Cybersecurity AI, focused on the transformative impact of AI and automation on cybersecurity strategies.

Other sessions at the conference discussed the revolutionary impact of AI on security and techniques for securing AI systems. The consensus was clear: deploying AI tools requires a strong approach to security, emphasizing trust boundaries and access controls.

NVIDIA at DEF CON 32

DEF CON 32, the world’s largest hacker conference, featured NVIDIA researchers supporting the AI Village. Attendees engaged in real-time hacking challenges focused on large language models (LLMs). The event showcased the potential of AI-powered tools to accelerate security research.

Adversarial Machine Learning Training

At Black Hat, NVIDIA and Dreadnode conducted a two-day training on machine learning (ML) to assess security risks against ML models and implement specific attacks. Participants gained hands-on experience critical for shaping effective defensive strategies.

Focus on LLM Security

NVIDIA Principal Security Architect Rich Harang presented on LLM security at Black Hat, emphasizing the importance of grounding LLM security in a familiar application security framework. Attendees were advised on maintaining robust security by identifying and analyzing trust and security boundaries.

Democratizing LLM Security Assessments

At DEF CON, NVIDIA introduced an open-source tool called garak for LLM security probing. This tool allows practitioners to test potential LLM exploits quickly, marking a significant step forward in standardizing security definitions for LLMs.

Summary

NVIDIA’s participation in Black Hat USA and DEF CON 32 highlighted its commitment to advancing AI security. The company’s contributions provided valuable knowledge to the security community. For more insights on NVIDIA’s work in AI and cybersecurity, visit the NVIDIA Technical Blog.

Image source: Shutterstock