ABOITIZ Data Innovation (ADI) proudly announced that its collaborative project with the UnionBank of the Philippines and the Smith School of Business at Queen’s University in Canada, which focuses on anti-discrimination law, artificial intelligence, and gender bias, has gained recognition for its significant contribution to protecting human rights.
The project was recently listed in the Global Top 100 Artificial Intelligence (AI) solutions by the International Research Centre on Artificial Intelligence (IRCAI) under the auspices of UNESCO.
“This recognition marks another significant milestone for our team. Our study emphasizes the importance of addressing gender bias in fintech lending and calls for rethinking existing anti-discrimination laws to ensure fairness and equality in the context of AI systems.,” Dr. David R. Hardoon, CEO of ADI said expressing his satisfaction with this achievement.
Recent studies have highlighted the inherent bias and inequality present in the consumer lending market, particularly affecting women and minority groups. Acknowledging these disparities, ADI conducted a study focused on non-mortgage fintech lending, employing explainable and responsible AI practices. The research aimed to determine whether excluding gender information from creditworthiness assessments would benefit or hinder the groups these measures are meant to protect. The study ultimately revealed that the exclusion of gender-related data significantly reduced gender discrimination and increased firm profitability.
The project serves as a guiding force in the improvement of anti-discrimination laws to ensure that Machine Learning (ML) models foster a fair and inclusive system, especially within the financial services industry, while simultaneously enhancing firm profitability.
Dr. Adrienne Heinrich, Head of ADI’s AI and Innovation Center of Excellence, expressed gratitude for being named one of the top 100 international AI solutions contributing to the 17 SDGs. She commented, “The outcomes of our study provide a solid foundation for deliberating on the use of protected attributes, such as gender data, in fintech, particularly credit lending models. Achieving a win-win situation for both businesses and communities ultimately depends on the responsible collection and use of gender data,
As AI systems become increasingly complex, achieving transparency, reliability, and interpretability poses challenges. In response, ADI has developed the Explainable and Responsible AI (XRAI) Guidelines to ensure ethical and appropriate use of AI.
These guidelines provide practical recommendations for leaders, team members, and stakeholders involved in the development and deployment of AI systems at ADI. They cover various topics, including data collection and management, model development and testing, human oversight and control, and the ethical implications of AI.
The guidelines are based on eight key principles: transparency, explainability, repeatability, safety and security, robustness, fairness, accountability, and human agency and oversight. By adhering to these principles, ADI aims to promote the reliable and responsible use of AI-driven technology, both within and beyond the organization, fostering trust in the technology and actively addressing concerns.
ADI’s commitment to the protection of human rights through AI innovation marks a significant stride forward. By championing fairness, equality, and inclusivity, ADI strives to create a future where AI serves as a tool for positive social impact, empowering individuals and communities alike.
“We take great pride in having our work acknowledged by IRCAI and UNESCO. With its profound implications for the betterment of society, we hope our findings contribute to the continued development of ethical and inclusive AI solutions,” Hardoon concluded.
IRCAI compiles the Top 100 list to highlight effective and impactful AI solutions that align with the 17 United Nations (UN) Sustainable Development Goals (SDGs). The project’s applicability, credibility, and ethical approach to solving SDG1: No Poverty, SDG5: Gender Equality, SDG10: Reduced Inequality, and SDG17: Partnerships to Achieve the Goal earned it recognition as an “excellent” research initiative, placing it within the top 30 percent of projects on the Global 100 list
