Request for Information from Industry – Countering Bias in AI/ML Datasets

This post benefited from significant contributions by Petruce Jean-Charles, Communications Associate for the Computing Community Consortium (CCC).

The U.S. Air Force Chief Scientist’s inter-agency working group on unintended AI bias is exploring the topic of bias in Artificial Intelligence (AI) and Machine Learning (ML) algorithms, with particular focus on datasets. They recently released a request for information (RFI) on countering bias in AI and ML datasets. As the landscape of AI expands, diverse domains present unique challenges concerning unintended bias. In this case, they seek insights from commercial and non-profit organizations harnessing data for AI development, encompassing both “narrow AI” and generative AI.

They are soliciting information to better understand:

  • The datasets and domain areas with prominent AI bias, including in government datasets
  • Ongoing efforts to address unintended bias in AI and ML datasets
  • Current partnerships between industry and government or academic institutions to study, identify, and mitigate AI bias
  • Existing limitations and obstacles to mitigating such bias
  • Areas of need for further research

Responses should be submitted through the portal by May 15, 2024 and should be a maximum of 5 pages.­ More information, including the full RFI, can be found here.

Academics intrigued by contributing their insights on these challenges are encouraged to engage with the group through a complementary Request for Information tailored for academia. They welcome responses from academic institutions, with a special emphasis on minority serving institutions (MSIs) and historically black colleges and universities (HBCUs).