Hackers Use AI in Pentagon-Backed AIxCC Cybersecurity Contest with Millions at Stake

Jul 28, 2024, 02:04 AM

Hackers are leveraging AI to combat cyberattacks in a Pentagon-backed competition, known as the AI Cyber Challenge (AIxCC). The contest, launched in 2023 and run by DARPA, aims to design AI programs that can scan open-source code for security flaws that bad actors might exploit. The competition has attracted significant attention due to the substantial monetary prizes at stake, with millions of dollars available for the winners. This event highlights the dual nature of AI in cybersecurity, as it can be used both to defend against and to perpetrate cyberattacks. Notable AI tools such as Evil-GPT and RogueGPT exemplify the potential and ethical risks involved. The Washington Post has provided detailed coverage of the latest strategies and technologies employed in this cutting-edge competition.