DoD AI Office Launches Bounty Program to Detect LLM Bias

The U.S. Department of Defense (DoD) Chief Digital and Artificial Intelligence Office (CDAO) has initiated the initial phase of the AI Bias Bounty program, a crowdsourced initiative designed to identify bias in AI systems.

It includes two exercises, the first of which is now open to the public and will run from  January 29 – February 27, 2024. Details on the second exercise have not been revealed yet.

“The goal of the first bounty exercise is specifically to identify unknown areas of risk in Large Language Models (LLMs), beginning with open source chatbots, so this work can support the thoughtful mitigation and control of such risks,” the DoD said.

The department is also encouraging active participation from the public. To participate, one does not require an extensive background in coding. Participants can also earn monetary bounties based on evaluation.

In November 2023, the DoD also unveiled a strategy aimed at expediting the integration of advanced AI capabilities. The goal is to secure enduring decision superiority for U.S. soldiers on the battlefield in the years ahead.

The new strategy, crafted by the Chief Digital and AI Office, expands on and surpasses the 2018 AI Strategy and the revised Data Strategy from 2020. These foundational documents have set the stage for the department's strategy in deploying AI-enabled capabilities.