CyberCast: Inside Oak Ridge National Lab's Pioneer Approach to AI

Edmon Begoli, Director of the Center for AI Security Research (CAISER) at the Energy Department's Oak Ridge National Laboratory, highlights how his team is ensuring safe and responsible AI deployment across government during the AI+ Expo in Washington, D.C., earlier this month. The research team is developing cutting-edge tools and methodologies to assess AI vulnerabilities, test models under extreme conditions and create energy efficient solutions for complex technological challenges. Begoli's team is focusing on two key innovations: comprehensive AI security assessments and TinyML, a revolutionary approach to developing small, energy-efficient machine learning models. By creating AI systems that can operate effectively in limited environments — such as surveillance drones or long-term sensor networks — the team is addressing both security concerns and environmental sustainability challenges facing the technology sector. The research goes beyond technical development, emphasizing education and public outreach to build AI literacy across diverse demographics. From high school students to senior citizens, the team is committed to demystifying AI technologies and demonstrating the potential to solve critical national security and infrastructure protection challenges. The team's work represents a proactive approach to technological innovation, ensuring that AI can be trusted, secure and strategically advantageous for the nation.