SeedAI, White House Jointly Announce Largest-Ever Generative AI Red Team
Announcement
SeedAI, White House Jointly Announce Largest-Ever Generative AI Red Team
By Austin Carson
May 4, 2023
Washington, D.C.

SeedAI, White House Jointly Announce the Largest-Ever Generative AI Red Team

SeedAI, The Wilson Center, and partners join the White House in announcing largest-ever AI Red Team exercise with AI Village @ DEF CON 31, hosting public officials and thousands of hackers & students

Today, SeedAI, the Wilson Center Science and Technology Innovation Program (STIP), AI Village, Humane Intelligence, and AVID – along with our community partners at Black Tech Street, Houston Community College, and the Congressional App Challenge – announced that we’re hosting thousands of hackers & students to test the AI models of Anthropic, Google, Hugging Face, NVIDIA, OpenAI, and Stability at DEF CON 31

SeedAI: 

“We’ve crossed into an AI-driven world – the winters are over. Current technological progress is incomparable in terms of speed, scale, and conveyance of individual power,” said Austin Carson, Founder & President of SeedAI. “We must build public spaces and tools such that people from all walks of life can trust & benefit from AI because they’re exploring, building, and testing it. This is the perfect program to allow all Americans to protect their interests while building necessary skills and understanding.”

The White House Office of Science and Technology Policy (OSTP):

“Testing and red teaming of AI models is an important component in effectively managing risks of the technologies. I applaud SeedAI, the AI Village, Humane Intelligence, AI Vulnerability Database, the Wilson Center Science and Technology Innovation Program, and their community partners Houston Community College and Black Tech Street for helping organize this effort, as well as the companies that have made independent commitments to participate in these public assessments,” said Arati Prabhakar, Assistant to President Biden and Director of the White House Office of Science and Technology Policy.

The Wilson Center STIP: 

“It’s rare that a technology delivers so much promise to revolutionize every aspect of our society, and it’s important that we do this right,” said Kellee Wicker, director of STIP at the Wilson Center. “We need to address short-term misinformation and bias risks, but also address long-term workforce effects by expanding AI skillsets for communities that are often left out of the early stages of significant technological change. I’m excited for the Wilson Center to partner with these incredible groups on an event that tackles both.”

Black Tech Street:

"Black Tech Street is beyond excited to be participating in the largest red team ever assembled for a group of A.I models. Artificial Intelligence may be the most transformative technology human beings have ever created, and it is critical that it is developed in a way that will exponentially accelerate the closing of inequity gaps rather than expanding them,” said Tyrance Billingsley II, Founder & Executive Director of Black Tech Street. “Black Tech Street is honored to serve on this Red team and look forward to keeping Black Wall Street at the forefront of technological innovation, where it will stay for the next 100 years."

Houston Community College:

"Houston Community College is thrilled to be part this initiative that is bringing talented students, teachers, industry professionals, and AI aficionados together to strengthen the security of large language models that are widely used around the world," said Dr. Madeline Burillo-Hopkins, HCC System Vice Chancellor for Workforce Instruction and HCC Southwest President.

AI Village @ DEF CON:

“Traditionally, companies have solved this problem with specialized red teams. However this work has largely happened in private,” said Sven Cattell, Founder of AI Village. “The diverse issues with these models will not be resolved until more people know how to red team and assess them. Bug bounties, live hacking events, and other standard community engagements in security can be modified for machine learning model based systems. These fill two needs with one deed, addressing the harms and growing the community of researchers that know how to help.” 

The National Science Foundation’s Computer and Information Science and Engineering (NSF CISE) Directorate and the National Institute for Standards and Technology (NIST) will also participate. The Congressional AI Caucus and Senate AI Caucus are collaborating on this initiative as part of their AI Primer series, and the exercise will be adapted into programming for policymakers – as well as for the national networks of our community partners.

###