White House launches AI-based contest to secure government systems from hacks

The White House has initiated a multi-million-dollar cyber competition aimed at promoting the application of artificial intelligence (AI) to detect and rectify security vulnerabilities within U.S. government infrastructure. This effort comes as malicious actors increasingly exploit AI for detrimental purposes in cyberattacks.

Anne Neuberger, the U.S. government’s deputy national security advisor for cyber and emerging technology, emphasized the urgency of cybersecurity, characterizing it as a contest between offensive and defensive measures. Neuberger noted that malevolent actors are already leveraging AI to expedite the identification of vulnerabilities and the development of malicious software.

Various U.S. entities, spanning healthcare organizations, manufacturing companies, government institutions, and more, have been targeted by cyberattacks in recent years. Officials have issued warnings about future threats, particularly from foreign adversaries.

The remarks made by Neuberger align with observations shared by Canada’s cybersecurity chief, Samy Khoury, who highlighted the diverse applications of AI in cyber activities, including crafting phishing emails, generating malicious code, and spreading disinformation.

The two-year cyber competition, boasting approximately $20 million in rewards, will be overseen by the Defense Advanced Research Projects Agency (DARPA), a U.S. government agency dedicated to developing technologies for national security. Notably, prominent technology companies such as Alphabet’s Google, Anthropic, Microsoft, and OpenAI are contributing their AI systems to address this challenge.

The competition serves as a notable step toward addressing an emerging threat that experts are still grappling to fully comprehend. Over the past year, several U.S. companies have introduced generative AI tools, like ChatGPT, enabling users to generate convincing videos, images, texts, and computer code. Similar models have also been launched by Chinese firms, indicating the potential for widespread misuse.

The DARPA AI challenge aspires to foster a community of cyber defenders who leverage the participating AI models to enhance cyber defenses, according to Neuberger. The Open Source Security Foundation (OpenSSF), a group of experts striving to enhance open source software security, will oversee the deployment of the “winning software code” to ensure immediate application, as outlined by the U.S. government.