Texas university kills 'biased' PhD applicant screening system
A machine learning-based algorithm used to evaluate applicants for the University of Texas at Austin ’s P hD in computer science was pulled after concerns about bias and inequality. Used from 2013, the computer science department's GRADE (GRaduate ADmissions Evaluator) algorithm was trained using the details of students acce p ted before 2013 in order to teach the system to identify people the school would favour. GRADE made the school's admissions process quicker, but was also found to have been disadvantaging underrepresented groups, notably women and Black people, on which the university was latterly placing greater emphasis. The system was terminated in December 2020 after a student backlash. Created by a University of Texas faculty member and graduate student, GRADE predicted how likely the admissions committee was to approve an applicant and expressed that prediction as a numerical score out of five. The system also explained what factors most impacted its decision. System 🤖 GRADE Operator: University of Texas at Austin Developer: U niversity of Texas at Austin Country: USA Sector: Education Purpose: Assess PhD applications Technology: Machine learning Issue: Bias/discrimination - race, gender Transparency : Governance
- Date it happened
- 2013-01-01
- Organisation involved
- University of Texas at Austin
- Product, system or model
- GRaduate ADmissions Evaluator (GRADE)
This incident was imported from AIAAIC and is used under CC BY-SA 4.0. Our additions to it — the structured fields, the translation, the checks against other reports — are published under the same licence.
This is a record of what was reported, not a finding that anyone broke the law. If it names your organisation and you believe it is wrong, the corrections process is free and open to everyone.