Webinar: Addressing the gender bias in artificial intelligence data
29 March 2021, 14:00 – 15:00 (Paris)
As part of the March on Gender, the OECD Directorate for Science, Technology and Innovation and the International Transport Forum (ITF) held a virtual event on gender bias in artificial intelligence (AI) data (March 29, 2021).
📹 👀 Watch the recorded version on Youtube:
This virtual event brought together technical and policy experts to address questions around gender bias in AI, such as:
- What are the best ways to combat gender bias in AI training data?
- How can data governance help to overcome challenges related to gender bias in data and other gender-related AI concerns?
- How can AI help to identify and eliminate gender bias?
- How can implementing the OECD AI Principles help to address gender bias?
Moderator: Audrey Plonk, Head of Digital Economy Policy Division, Directorate for Science, Technology and Innovation, OECD
Opening remarks by:
Mr. Ulrik Vestergaard Knudsen, Deputy Secretary-General, OECD
Dr. Young Tae Kim, Secretary-General of the International Transport Forum
Panel discussion and live Q&A with:
Dr. Antonella Santuccione Chadha, Co-founder and CEO of the Women’s Brain Project (WBP), and Head of Stakeholder Engagement at Biogen
Mr. Lee Glazier, Head of Service Integrity, Civil Aerospace, Rolls-Royce Plc
Dr. Aleksandra Mojsilović, IBM Fellow, Head of AI Foundations at IBM Research, and Co-Director of IBM Science for Social Good
Ms. Cristina Pombo, Principal Advisor and Head of the Digital and Data Cluster, Social Sector, Inter-American Development Bank
Dr. Jeni Tennison, Vice President and Chief Strategy Adviser, Open Data Institute and Co-Chair, Working Group on Data Governance – The Global Partnership on AI (GPAI)
Data is the lifeblood of the digital economy and artificial intelligence, so if data processes yield data that are non-representative then the resulting algorithms will almost certainly have biased outcomes. Algorithms that do not take pre-existing gender-based inequalities into account will also be biased.
While AI systems run the risk of perpetuating or even exacerbating gender-based inequalities, AI can also help to identify and correct human bias.
Background documents
- Artificial Intelligence in Society, OECD (2019),
- Trustworthy AI in health. Background paper for the G20 AI Dialogue, Digital Economy Task Force (April 2020)
- The role of Data in AI. Report for the Data Governance Working Group of the Global Partnership of AI. Report prepared for GPAI by the Digital Curation Centre, Edinburgh University School of Informatics and Trilateral Research (2020),
- AI Fairness 360: An extensible toolkit for detecting, understanding, and mitigating unwanted algorithmic bias, Rachel K. E. Bellamy, et al. (2018),
- Sex and gender differences and biases in artificial intelligence for biomedicine and healthcare, Davide Cirillo, et al., NPJ Digital Medicine 3.1 (2020).