Agentic 101 - Washington, DC (State of the Stack) · · Washington, DC ·
Red-teaming for Agentic Safety and Security
Abstract
Adrianna will talk about how to get started with agentic red-teaming: - how to implement guardrails and prove (or disprove) that they work - how to prevent your bot from diverging from the approved workflow - how to collect and present evidence for auditors
Speaker
Adrianna Tan
Adrianna is the founder of Future Ethics Lab, and she focuses on AI evaluations, red teaming, deployment readiness, and governance, particularly in high-stakes environments such as government, healthcare, and public services. Over the past several years Adrianna has worked with governments, researchers, and private-sector organizations to understand how AI systems behave under real-world conditions, including multilingual and multicultural contexts that are often overlooked in standard benchmarks and testing approaches.