The White House unveiled a multi-million-dollar cybersecurity competition aimed at fostering the utilization of artificial intelligence (AI) to detect and rectify security vulnerabilities within the infrastructure of the U.S. government.
This initiative comes in response to the increasing adoption of AI by hackers for malicious endeavors.
Anne Neuberger, the U.S. government’s Deputy National Security Advisor for Cyber and Emerging Technology, remarked, “The realm of cybersecurity is a constant battle between offensive and defensive actions.” She further explained, “We are aware that malicious entities are already leveraging AI to expedite the identification of vulnerabilities and the creation of harmful software.”
Various American entities, ranging from healthcare organizations to manufacturing companies and governmental institutions, have been subjected to cyberattacks in recent times. Officials have also cautioned against potential future threats, particularly from foreign adversaries.
Neuberger’s observations about AI align with the sentiments expressed by Canada’s Chief of Cybersecurity, Samy Khoury, last month. Khoury noted that his agency had observed AI being employed for tasks encompassing the crafting of phishing emails, the composition of malicious code, and the dissemination of misinformation.
Two years
The competition, spanning two years, offers approximately $20 million in rewards and will be overseen by the Defense Advanced Research Projects Agency (DARPA), the governmental body responsible for the development of technologies pertaining to national security, according to a statement from the White House.
Key players in the AI landscape, including Alphabet’s (GOOGL.O) Google, Anthropic, Microsoft (MSFT.O), and OpenAI, which are at the forefront of the AI revolution in the U.S., have committed to making their AI systems available for the challenge, as per government sources.
This competition reflects official efforts to address an emerging threat that experts are still in the process of comprehending fully. Over the past year, numerous U.S. entities have introduced generative AI tools, like ChatGPT, that enable users to generate convincing videos, images, texts, and code. Similar models have been introduced by Chinese companies to catch up in this domain.
According to experts, such tools could substantially simplify tasks such as conducting large-scale hacking campaigns or creating fabricated profiles on social media platforms to propagate false information and propaganda.
Neuberger emphasized the objective of the DARPA AI competition, stating, “Our aim through this challenge is to stimulate a broader community of cybersecurity defenders who can leverage the participating AI models to enhance our cyber defenses using generative AI technologies.”
The Open Source Security Foundation (OpenSSF), a U.S. consortium of experts dedicated to enhancing the security of open-source software, will be responsible for ensuring the “winning software code is promptly implemented,” as indicated by the U.S. government.
