Lee Lounge/109

In our session, we will:

1. Provide an overview of present-day AI capabilities and the risks and harms stemming from their increasingly widespread adoption, and give an approachable introduction to concepts in AI Safety, Ethics, and Security.
2. Present interactive, live demonstrations recently shown to Congressional Staffers in D.C. at the Center for AI Policy’s Advanced AI Expo. We will demonstrate the security risk of voice models that mimic emotional speech—posing threats to businesses, enabling floods of fake 911 calls, or misleading Congressional representatives through coordinated fake constituent campaigns. We will also show how integrating LLM agents into business workflows can create unique vulnerabilities in multi-agent systems.
3. Demonstrate how, even at current capability levels, AI systems already pose significant risks relevant to IT professionals.
4. Outline how AI capabilities have developed over recent years and present arguments for why this rapid pace will continue. We’ll briefly explore the implications for industry and society, including risks associated with frontier and future AI, and provide a high-level overview of mitigation efforts underway.

We will close with a discussion of our work at WAISI, how to get involved in responsible AI, and what this means for IT professionals.

Presenters

• Lanya Yu
• William Anderson
• Kaushal Dasika
• Alka Lakadia
• Viktoria Sakman

Recording (with captions)

Recording (with audio description)