August 10, 2023 at 10:30 GMTModified date: August 10, 2023 at 10:30 GMT
August 10, 2023 at 10:30 GMT

US launches AI Cyber Challenge ‘to protect Americans’

The Biden-Harris administration has announced a new artificial intelligence competition as it seeks to utilise AI to protect its most important software.

US launches AI Cyber Challenge ‘to protect Americans’

The Biden-Harris administration has announced the launch of a new competition revolving around artificial intelligence (AI) as it seeks to utilise AI to protect the United States’ most important software.

Called the ‘AI Cyber Challenge’ (AIxCC), the contest will run for two years and will feature almost $20million in prizes.

The challenge is led by the Defense Advanced Research Projects Agency (DARPA). It will bring together several leading AI companies, like Anthropic, Google, Microsoft, and OpenAI, in a partnership with DARPA to identify and fix software vulnerabilities using the novel technology.

Through this contest, the government aims to drive the creation of new technologies to rapidly improve the security of computer code that helps run the internet and its critical infrastructure. The competition’s official website stated: “Competitors will leverage advances in AI to invent the next generation of cybersecurity defences for today’s digital society. Over the next two years, AIxCC will challenge teams to develop AI-drive systems to automatically secure the critical code that underpins daily life.”

The leading AI companies in the partnership will provide their powerful and cutting-edge technology to the competitors who will use it in designing new cybersecurity solutions. A total of $7m has also been set aside to be granted to small businesses wishing to compete. This is to ensure a broader participation and a level playing field.

The qualifying event will take place in Spring 2024 where the 20 top scoring teams will be invited to participate in the semi-final competition at DEF CON 2024, a major cybersecurity conference. Out of this, five top scoring teams will receive monetary prizes and will be qualified for the final phase of the competition which will be held at DEF CON 2025. Here, the top three scoring competitors will receive additional monetary prizes.

AIxCC will open two tracks for participation: an Open Track and Funded Track. While Open track will be available for everyone, the latter will be reserved for seven small businesses who will be funded by DARPA. The proposal to be considered for the same has to be submitted between 17 August and 19 September 2023, whereas the Open Track will have its registration from 1 November and 15 December 2023.

The winner of the competition will bag a $4m worth cash prize. Second position will be given $3m and $1.5m will be granted to the team coming third. Participants of the contest will have to publicly share the inner workings of their systems, so that a broader utilisation of their solutions can be achieved.

The Biden administration had first announced AIxCC at the nation’s largest hacking conference called the Black Hat USA Conference in Las Vegas, Nevada. The role of the challenge advisor in providing guidance to the teams has been taken up by the Open Source Security Foundation (OpenSSF), a project of the Linux Foundation.

This is not the first time the US government has fostered innovation through competitions. The year 2014 saw DARPA initiating the Cyber Grand Challenge, which aimed at creating an open-source automated defence system capable of safeguarding computers against cyber threats.

The latest competition is a part of its broader commitment towards harnessing the powers of AI to address the nation’s challenges. The administration had announced a commitment from several AI companies to participate in an independent, public evaluation of large language models (LLMs) at DEF CON 2023 earlier this year. This was also aimed to help advance safer, more secure and more transparent AI development.

On the legislative front, the Biden-Harris Administration is currently developing an executive order in curating bipartisan legislation to “help America lead the way in responsible AI innovation”.