The Defense Innovation Unit publishes ‘Responsible AI Guidelines’> US Department of Defense> Defense Department News

The Defense Innovation Unit released its first “Responsible AI Guidelines” document on November 15 with the aim of operationalizing the Department of Defense’s ethical principles for artificial intelligence in its commercial prototyping and procurement efforts.

The DIU’s RAI Guidelines provide a step-by-step framework for AI companies, DOD stakeholders and program managers that can help ensure that AI programs are built on the principles of fairness, accountability and transparency at every step of the the development cycle of an AI system, “said Jared Dunnmon, PhD, Technical Director of Artificial Intelligence / Machine Learning Portfolio at DIU.

The DIU team has spent the past 18 months working with researchers at the Carnegie Mellon University Software Engineering Institute and talking to industry partners, the Joint Artificial Intelligence Center, academia and officials and testing these guidelines to request useful feedback, Dunnmon said. They are intended specifically for use on DIU programs.

The goal of the guidelines, he said, is to:

  • Accelerate programs from the outset by clarifying end goals, aligning expectations, and recognizing risks and trade-offs.
  • Increase confidence that AI systems are developed, tested and controlled to the highest standards of fairness, accountability and transparency.
  • Support changes in the way AI technologies are evaluated, selected, prototyped, and adopted to avoid potential adverse outcomes.
  • Induce questions and conversations that are crucial to the success of the AI ​​project.

The guidelines provide examples of how responsible AI considerations can be put into practice in real-world applications in an effort to create a user-friendly and easier-to-understand document that speeds up the process, Dunnmon said.

“Users want to be able to trust and verify that their tools protect US interests without compromising our collective values,” said John Stockton, co-founder of Quantifind, a software technology company that provided DIU feedback on the guidelines under their prototype project. “These guidelines show promise in actually accelerating technology adoption, as it helps identify and address potential show-stopping issues. We have found that leaning into this effort has also served us well outside the government by to strengthen internal controls and create transparency and patterns of trust that can also be exploited by all users, both public and private. “

To view the guidelines, visit:

Leave a Comment