skip to main content
Article Podcast Report Summary Quick Look Quick Look Video Newsfeed triangle plus sign dropdown arrow Case Study All Search Facebook LinkedIn YouTube Bluesky Threads Instagram Right Arrow Press Release External Report

Reflections from Elizabeth Adams:

  • Adams emphasized the importance of engaging proactively in new technologies. She noted that proactive engagement had the potential both to prevent technophobia and to ensure technology is adopted effectively and ethically. Taking action before tools are implemented increases the likelihood of community buy-in and (hopefully) decreases the likelihood that the tool will be abused.
  • One theme was how users can know whether a tool is unbiased. Adams shared that it helps to know that a company has an AI ethics framework in place, and that it is important to ask about the data that are being used to train AI systems. She acknowledged that there are limits to what the public can do if a company is not forthcoming, but also noted that it was possible to make the inclusion of an AI ethics policy a requirement for winning a contract.
  • Adams pointed out that it is not necessary to be a software engineer or contract administrator to be a part of this discussion. She noted that those with non-technical and non-development backgrounds could set up a monthly check-in meeting with these team members or develop a standard list of questions to ask developers. There are also resources online to help educate non-technical leaders and better enable them to push for responsible and equitable AI. Contract administrators can make AI ethics guidelines a requirement for securing work, program managers can insist on regular discussions about ethical issues that might arise from AI work, and analysts can ask questions about the source of data used to train AI systems. We all have a role to play.
  • Finally, we must change the narrative of who belongs at the table. Members of groups affected by AI bias have perspectives that are invaluable to discussions about AI. Welcoming these perspectives might require looking outside one’s own organization and tapping into organizations that are engaged in grassroots equity work. This can be challenging in a national security setting because of security clearance issues, but it is nevertheless critical that we connect with local populations and ensure that these perspectives are heard as frequently as possible.
  • Many of the challenges the Department of Defense faces as it looks to integrate AI into national security work are the same challenges that companies face in the private sector: What are my data? Where are my data coming from? Who curated my dataset? Who are my partners? What is my role in the solution?
Download report

Distribution unlimited.

Details

  • Pages: 1
  • Document Number: CCP-2022-U-032876-Final
  • Publication Date: 6/23/2022
Back to Navy Human Resources