Biden administration launches AI cybersecurity challenge to ‘protect Americans’

With an allocation of nearly $20 million in rewards, the AI Cyber Challenge brings together prominent AI enterprises such as Anthropic, Google, Microsoft, and OpenAI.

The Biden administration revealed an opportunity on August 9, for hackers to vie for substantial monetary rewards through the application of artificial intelligence (AI) in safeguarding vital United States infrastructure from cybersecurity vulnerabilities.

In spring, a preliminary phase will select up to 20 high-performing teams for DEF CON 2024’s semifinals. Of these, a maximum of five teams will earn $2 million each and move on to DEF CON 2025’s finals. The top three teams will vie for extra prizes, including a $4 million award for the best safeguarding of vital software, as stated in an official press release.

With an allocation of nearly $20 million in rewards, the AI Cyber Challenge brings together prominent AI enterprises such as Anthropic, Google, Microsoft, and OpenAI. These industry leaders will contribute their technology to the competition, which was unveiled during the Black Hat U.S.A. hacking conference held in Las Vegas.

Participants will be requested to publicly share the inner workings of their systems, enabling broader utilization of their solutions. Additionally, guidance for the challenge is provided by the Open Source Security Foundation, a division of the Linux Foundation.

The organizing body of the competition, the Defense Advanced Research Projects Agency (DARPA), has committed to offering financial support of up to $1 million to seven small enterprises aiming to join the competition, thus ensuring a diverse range of participants.

Related:US House Financial Services Committee leadership takes opposing sides on the PayPal stablecoin

The use of hacking competitions to foster innovation is not a new approach for the government. Back in 2014, DARPA initiated the Cyber Grand Challenge, aimed at creating an open-source automated defense system capable of safeguarding computers against cyber threats. The present two-year challenge follows a comparable framework to this prior initiative.

The contest shows that there are official efforts to deal with a new threat that experts are working to fully understand. In the past year, several U.S. companies have created different AI tools, like ChatGPT, that let users make realistic videos, images, texts and code.

Magazine: Experts want to give AI human ‘souls’ so they don’t kill us all