DARPA Announces $4M AI Cyber Challenge Winners in Security Showdown
▼ Summary
– Team Atlanta won the AI Cybersecurity Challenge (AIxCC) at DEFCON 33, securing a $4m prize as a collaboration between Georgia Tech, Samsung Research, and Korean institutions.
– Trail of Bits and Theori placed second and third, winning $3m and $1.5m respectively, in DARPA’s competition to develop AI-powered cybersecurity tools.
– The top three teams’ cyber reasoning systems were open-sourced, with three additional models set to be released soon, as announced by DARPA’s director.
– AIxCC, launched in 2023, aimed to create AI tools for securing US critical infrastructure, with $2m awarded to each of the seven finalists at DEFCON 32.
– The competition was backed by tech giants like Google and Microsoft, with DARPA and ARPA-H adding $1.4m to the original $29.5m prize pool.
The winners of DARPA’s groundbreaking AI Cyber Challenge (AIxCC) have been crowned after an intense two-year competition aimed at revolutionizing cybersecurity through artificial intelligence. Team Atlanta emerged victorious, taking home the $4 million grand prize at DEFCON 33. This elite group brought together top minds from Georgia Tech, Samsung Research, and leading South Korean institutions to develop cutting-edge AI solutions for critical infrastructure protection.
New York cybersecurity firm Trail of Bits secured second place with a $3 million award, while Theori, a collaborative team spanning the U.S. and South Korea, claimed third with $1.5 million. All three teams contributed to open-source cyber reasoning systems now available for public use, with additional models set for release in coming weeks according to DARPA director Stephen Winchell.
Launched at Black Hat 2023, the AIxCC initiative challenged experts to create AI-driven tools capable of identifying and patching software vulnerabilities faster than ever before. Backed by DARPA and ARPA-H, the competition attracted seven finalists who each received $2 million in development funding. Major tech players including Google, Microsoft, and OpenAI supported the event with over $1 million each in computational resources.
The U.S. government has committed significant resources to the program, with Deputy HHS Secretary Jim O’Neill announcing an additional $1.4 million investment beyond the original $29.5 million prize pool. This underscores the growing importance of AI-powered cybersecurity in protecting national infrastructure and healthcare systems against evolving digital threats.
With winning solutions now publicly accessible, the AIxCC marks a major step toward automated vulnerability detection and response, potentially transforming how organizations defend against cyberattacks in real time.
(Source: InfoSecurity)
