As artificial intelligence (AI) becomes more widely used, more examples of inequitable design, usage, and effect of AI-enabled products in unjust developing nations are emerging.
Because humans unknowingly integrate prejudice in AI technology during data collecting, model creation, or end-use applications, the majority of these approaches and techniques have the potential for inequitable consequences across genders.
As such, these technologies often put those that are marginalized in low- and middle-income nations in danger and lost chances.
More measures are necessary to discover unique and creative techniques to support decision-makers overcome existing and potentially discriminatory practices and inequitable results emerging from AI technology in order to build an equitable and fair digital ecosystem.
Via Digital Frontiers, the United States Agency for International Development (USAID) is sponsoring a challenge that will award one or more grants for novel innovative approaches to identifying and addressing existing and prospective gender inequalities in AI systems important to world development situations.
USAID wants to help develop strategies to improve AI system prevention, detection, transparency, monitoring, and accountability so that their outputs don’t cause gender-inequitable outcomes.
Technology-based solutions may or may not be used. With a total of $375,000 in award cash, Digital Frontiers expects to disburse up to five prizes. The task will be split into two parts.
A call for short ideas notes is being issued. Concepts will be welcomed into one of two paths: “Prevention,” which will focus on ways to avoid AI systems from integrating gender bias into their outputs, and “Response,” which will focus on ways to increase openness and accountability in current AI systems.
Semi-finalists selected from Phase One will be welcomed to a co-creation workshop in early 2022 to examine the environment in which examples of unfairness develop, exhibit potential remedies, and encourage cooperation and partnership among semi-finalists within and across tracks.
Semi-finalists will be asked to submit amended idea notes, which will reflect any changes to their initial concept note.
Grant funding will be provided to winning applicants to create, test, scale, and/or socialize the strategy over a period of a year.
Proposals from the following groups are encouraged by Digital Frontiers:
- Technology companies, companies, small and medium businesses, civil society organizations, and researchers from throughout the globe, no matter the size.
- Different teams have clear, planned, and collaboration solutions or frameworks to address the complexities of gender imbalance and unfairness that arise as a result of AI applications, including partnerships between developers, AI tool customers, and civil society organizations.
- Companies who offer a business case for why tackling gender inequitable results from AI is both beneficial to their organization and sustainable after the grant has been awarded.
- Proposals that show developers are making a good faith attempt to lead the industry in identifying and eliminating instances of gender injustice in their AI technologies.
- Proposals that take into account the range of AI-enabled bias and unfairness that women encounter based on their intersectional identities which include disability, sexual minority, language minority, caste status, etc.
How to apply
Do you have an interest in addressing gender inequlaity with AI technology? Apply here. The deadline for application is Thursday, January 6, 2022.
Don’t miss important articles during the week. Subscribe to techbuild weekly digest for updates