How to Reduce AI Bias in Machine Learning Models: Strategies for Fairness, Accountability, and Transparency to Improve Business Outcomes.
This presentation will explore how AI systems can inadvertently perpetuate societal and historical biases. Drawing upon interdisciplinary research from law, sociology, and psychology, the presentation will discuss how machine learning (ML) models often inherit biases from their training data, with a special focus on Canadian institutions leading the way in "explainability," "fairness," and "responsible AI." Using research, and emerging best practices, attendees will gain an understanding of the challenges and opportunities surrounding AI bias. Finally, the presentation will discuss key risk areas for organizations to considerations by highlighting digital transformation risks, operational risks, and AI project risks, including risk management strategies and implementation challenges.
Learning Objectives:
- Learn the origins of bias in AI systems, particularly in the context of ML models
- Learn how algorithmic bias can perpetuate societal inequities
- Learn how to apply emerging best practices and strategies for promoting fairness, accountability, and transparency in AI implementation projects
- Learn how to use practical frameworks and tools for assessing and mitigating bias in real-world application of ML models
- Learn how the critical role of a diverse group of stakeholders can help build ethically responsible and inclusive AI systems
Speaker: Pulkit Mogra, University of Ottawa
Pulkit Mogra specializes in the intersection of privacy law, AI regulation, and cybersecurity, with a focus on making complex technologies safer and more equitable. He is a Ph.D. candidate at the University of Ottawa, where his research on predictive algorithms examines how AI affects different communities and what safeguards are needed to prevent bias has relevant application for most organizations as they balance between technological innovation and ethical considerations. Pulkit also consults with stakeholders to evaluate their AI systems for bias, to explain the legal implications of AI-generated products, and to help them navigate the complex landscape of technology regulation. Pulkit has an LL.M. in Law and Technology from Tel Aviv University and taught at the University of Petroleum and Energy Studies in India. You can contact Pulkit at pmogr089@uottawa.ca or www.linkedin.com/in/pulkitmogra
This Community Conversation is brought to you by AIIM's True North Special Interest Group
This webinar is eligible for 1 CEU under Domain 4.
#AIIMConference
#ArtificialIntelligence
#DigitalTransformation
#InformationGovernance
#InformationManagement
#MachineLearning
#RiskManagement
#AIBias
#AIEthics