Experience the difference. Meet your peers, talk with experts, build your network, and turn insights into action.
Christopher S. Campbell
Director of AI Governance
Lenovo
Speaker Bio
Christopher serves as both Director of AI Governance and Global Product and Services Security Leader within the Chief AI Office at Lenovo. He oversees the Lenovo Responsible AI program and delivers strategic initiatives for secure development and deployment of AI systems, as well as risk management and monitoring for AI regulatory compliance. He has over twenty years of global experience at the intersection of human safety and cyber and product security, including connected vehicles, surgical robotics and healthcare. Christopher has a Bachelor of Science in Cognitive Psychology, an MBA and a CISSP.
Show More
Show Less
Session
Monday, March 09, 2026 04:30 PM - 05:00 PM EDT
Executive Story: AI Risk from Tennessee to Timbuktu: A Human-Centric Approach for LLM Risk Management and Explainability
Christopher S. Campbell,
Director of AI Governance, Lenovo
“But what does that really mean?” is always the first question after reviewing an LLM security and safety test report.
The real question is how do we equate qualitative and subjective AI LLM risks, such as bias and toxicity, into quantitative risks that impact the business? Throw out the arbitrary scoring and graphs of traditional model testing reports and show the real impact to both humans and to the organization’s bottom line.
In this session, we will discuss a novel approach to evaluating and mitigating AI model security and privacy risks, in which we take a complex technology that is inherently designed to mimic a human and evaluate its behavior and risk like it is a human. This enables continuous governance and compliance, as well as true visibility into the dynamic nature of AI trust and safety at a region-by-region level. ... Show More Show Less
The real question is how do we equate qualitative and subjective AI LLM risks, such as bias and toxicity, into quantitative risks that impact the business? Throw out the arbitrary scoring and graphs of traditional model testing reports and show the real impact to both humans and to the organization’s bottom line.
In this session, we will discuss a novel approach to evaluating and mitigating AI model security and privacy risks, in which we take a complex technology that is inherently designed to mimic a human and evaluate its behavior and risk like it is a human. This enables continuous governance and compliance, as well as true visibility into the dynamic nature of AI trust and safety at a region-by-region level. ... Show More Show Less