Washington, D.C. — In a significant move to bolster cybersecurity and protect critical software infrastructure, the Biden-Harris Administration has unveiled the “AI Cyber Challenge” (AIxCC), a comprehensive two-year competition that harnesses the power of artificial intelligence (AI) to identify and address vulnerabilities in the United States’ most vital software systems, including those that power the internet and critical infrastructure. The initiative, spearheaded by the Defense Advanced Research Projects Agency (DARPA), aims to collaborate with leading AI companies, including Anthropic, Google, Microsoft, and OpenAI, to advance the security of computer code and contribute to the responsible advancement of emerging technologies.
The AI Cyber Challenge will present competitors from across the nation with the task of leveraging AI to identify and rectify software vulnerabilities. DARPA, in collaboration with the aforementioned top AI firms, will provide cutting-edge technology and expertise, thereby facilitating the development of new tools and strategies to enhance the security of computer code, addressing a pressing concern in the realm of cybersecurity.
The competition boasts a total prize pool of almost $20 million, making it a compelling incentive for participants. Small businesses interested in the competition will also be offered support through a $7 million allocation from DARPA, ensuring a diverse range of participants and a level playing field.
The competition will unfold in several stages. In Spring 2024, a qualifying event will take place, wherein up to 20 top-performing teams will be selected to move forward to the semifinal competition at DEF CON 2024, a renowned cybersecurity conference. Among these, up to five teams will receive monetary rewards of $2 million each and advance to the final phase, scheduled for DEF CON 2025. The three teams with the highest scores in the final competition will secure additional monetary prizes.
The AI Cyber Challenge announcement took place at the Black Hat USA Conference in Las Vegas, a prestigious hacking conference with a history of driving cybersecurity innovations. By adopting an automated and scalable approach to identifying and resolving vulnerabilities, the competition aligns with the conference’s legacy.
The Open Source Security Foundation (OpenSSF), a project under the Linux Foundation, will serve as a challenge advisor, facilitating the practical implementation of winning software solutions to protect crucial software infrastructure.
This initiative underscores the Biden-Harris Administration’s commitment to leveraging AI for the greater good and responsible technological advancement. It follows recent efforts by the administration, including securing commitments from leading AI companies to manage risks associated with AI technologies and facilitating an independent evaluation of large language models for enhanced transparency and security.
President Biden’s executive order, along with ongoing bipartisan legislation efforts, demonstrates the administration’s dedication to fostering responsible innovation in the field of AI and ensuring the security and welfare of the American people in an increasingly digital world.
