Texas Implements AI Grading of STAAR Exams, Sparking Concerns Over Fairness and School Accountability
• Texas is using computers to automatically grade open-ended STAAR exam questions this year to save $15-20 million per year on human graders.
• The grading engine uses artificial intelligence technology and was trained on a sample of 3,000 human-scored responses.
• Educators and parents worry about trusting computers to grade high stakes tests fairly and accurately.
• STAAR test performance impacts school accountability grades and can lead to state takeover of failing schools.
• With computer grading now implemented, districts are focused on preparing students to perform well for the automated scoring system.