Harini Kanakala

This talk explores how red teaming helps uncover vulnerabilities in large language models (LLMs) through adversarial testing techniques.
It also introduces the emerging risks of agentic AI systems and the evolving challenges they pose for secure AI development.

Biography:

Dr. Nagadhara Harini Kanakala is an AI researcher and red team specialist with deep expertise in adversarial testing, trustworthy AI, and the secure deployment of large language models (LLMs).

She leads advanced red teaming initiatives to uncover critical vulnerabilities across frontier models and guide organizations in evaluating real-world model behavior under adversarial conditions.

Harini has designed and deployed secure AI workflows across various cloud services, and actively contributes to building AI governance frameworks aligned with the NIST AI Risk Management Framework and the EU AI Act. Her work blends model development, offensive testing, and compliance documentation to support the safe and responsible scaling of AI systems.

Email:

Address:United States