View Proposal

Proposer
Hind Zantout
Title
Detecting Bias in Generative AI
Goal
To investigate bias in generative AI
Description
Harvard University have compiled an Implicit Attitude Test https://implicit.harvard.edu/implicit/takeatest.html that can be used to measure bias. Working with Dr. Hajar Yekani in SoSS, this project will investigate bias in generative AI systems. It could be using text or images. Interested students should first read up on the test and the different types of tests that are available. Then a suitable framework using genrative AI can be suggested to measure bias in such systems.
Resources
Background
Url
Difficulty Level
Moderate
Ethical Approval
InterfaceOnly
Number Of Students
2
Supervisor
Hind Zantout
Keywords
Degrees
Master of Science in Artificial Intelligence
Master of Science in Data Science
Back to List