As the use of artificial intelligence (AI) proliferates, more instances of the inequitable design, use, and impact of AI-enabled tools in developing countries are coming to light. Many of these tools and approaches have the capacity for inequitable outcomes across genders due to humans unconsciously embedding bias in AI technology during data collection, model design or end-use applications. As a result, these tools often pose the greatest risk of harm and missed opportunities to those already marginalized in low- and middle-income countries. To foster an equitable and inclusive digital ecosystem, more efforts are needed to identify innovative and creative approaches to help decision-makers address actual and potential gender biases and inequitable outcomes resulting from AI technology.

RevisedAIGraphic (002)-e856fd.png

Equitable AI Challenge: Sourcing Approaches to Prevent and Respond to Gender Inequity in AI

The U.S. Agency for International Development (USAID), through Digital Frontiers, is launching a challenge leading to one or more grants for innovative approaches to help identify and address actual and potential gender biases in AI systems relevant in global development contexts. USAID seeks to support approaches to increase the prevention, identification, transparency, monitoring and accountability of AI systems so that their outputs do not produce gender-inequitable results. Approaches may or may not be technology-based. Digital Frontiers anticipates disbursing up to five grants using a total of $375,000 in award funding. The challenge will take place over two phases:

  • Phase One: This is an open call for short concept notes. Concepts will be accepted under two tracks—one track called “Prevention,” for approaches to prevent AI systems from incorporating gender bias into their outputs, and one called “Response,” for approaches to inspire transparency of and accountability in existing AI systems. Strong applications from Phase One will be designated semi-finalists and invited to attend a co-creation workshop in early 2022 to discuss the context in which different instances of inequity arise, showcase different interventions, and encourage collaboration and partnership among semi-finalists within and across the tracks.
  • Phase Two: Semi-finalists will then be invited to submit revised concept notes that reflect updates to their original concept note as relevant. Winning submissions will be awarded grant funds to design, test, scale, and/or socialize the approach as appropriate over the course of one year.

Digital Frontiers encourages proposals from:

  • All organizations regardless of size; technology firms, startups, small, and medium enterprises, civil society organizations, and researchers from around the world.
  • Diverse groups that have clear, strategic, and collaborative approaches or frameworks to tackle the complexity of the gender bias and inequity occurring as a result of AI applications, including and especially, collaboration between developers, AI tool buyers, and civil society organizations.
  • Organizations that present a business case for why addressing gender inequitable outcomes from AI is both valuable to their organization and will be sustained following the grant.
  • Proposals that represent a good faith effort from developers to serve as an industry example on the steps needed to identify and eliminate instances of gender inequity in their AI tools.
  • Proposals that incorporate the diversity of AI-enabled bias or inequity women face according to intersectional identities (disability, sexual minority, language minority, caste status, etc.).

Apply Now

Deadline for Concept Notes is January 6, 2022, by 5pm EST. A link to RFP can be found here.