About the Talk:

Red teaming is an integral part of OpenAI's iterative deployment process. Over the past few years, our red teaming efforts have grown from a focus on internal adversarial testing at OpenAI, to working with a cohort of external experts to help develop domain specific taxonomies of risk and evaluating possibly harmful capabilities in new systems. This talk will explore the application of red teaming methodologies to assess the vulnerabilities and limitations of models such as GPT-4 and DALL-E 3.

Read the Blog, Lessons Learned on Language Model Safety and Misuse here.

About the Speaker:

Lama Ahmad is on the Policy Research team at OpenAI, where she leads efforts on external assessments of the impacts of AI systems on society. Her work includes leading the Researcher Access Program, OpenAI's red teaming efforts, third party assessment and auditing, as well as public input projects.

About the Host:

Natalie Cone leads OpenAI’s interdisciplinary community, the Forum, a community designed to unite thoughtful contributors from a diverse array of backgrounds, skill sets, and domain expertise to enable discourse related to the intersection of AI and an array of academic, professional, and societal domains. Before joining OpenAI, Natalie managed and stewarded Scale’s ML/AI community of practice. She has a background in the Arts, with a degree in History of Art from UC, Berkeley, and has served as Director of Operations and Programs, as well as on the board of directors for the radical performing arts center, CounterPulse, and led visitor experience at Yerba Buena Center for the Arts.