Back to resources

What students at UC Berkeley say about AI for assessment

Initial findings from student interviews at UC Berkeley

Published on 10/28/2025
1 min read
A one-day visit to UC Berkeley involved asking 115 students how they feel about teachers using AI to assist with grading or feedback. The goal was to understand whether students are comfortable with it and under what conditions. Preliminary findings show a mix of reactions: some said yes right away, others said no, and some said yes only if certain conditions are met. Among the latter, students mentioned conditions such as "as long as the teacher still reviews it," "as long as it's at least as good as a human," and "as long as I know when AI is being used." For a summary of responses, see Table 1. A faculty-level breakdown and an analysis of the most frequently cited conditions will follow soon.

Preliminary Results

Table 1: Distribution of responses from 115 UC Berkeley students.

Response typePercentageDescription
Yes21,7%Comfortable with AI in grading
Yes, only if57,4%Acceptable only under certain conditions
No20,9%Prefer teachers not to use AI

Request access

Assessment support for teachers
in the AI era