From Awareness to Action? The Impact of CO2 Emission Feedback on Student LLM Usage

MCML Authors

Ailin Liu

→ Group Frauke Kreuter
 Social Data Science and AI

Christoph Kern

Prof. Dr.

Core PI

Social Data Science and AI

Frauke Kreuter

Prof. Dr.

Core PI

Social Data Science and AI

Abstract

Given the growing adoption of Large Language Models (LLMs), mitigating the environmental impacts of AI-user interactions becomes a critical concern. We study whether real-time feedback on resource consumption changes LLM usage patterns and reduces CO2 emissions in an educational setting. Our use case draws on a between-subject experimental design and rich AI-user interaction data collected on an AI learning platform. Our results show lower emissions and total tokens per prompt for the real-time feedback group, indicating that informational feedback can influence selected measures of students’ LLM usage.

inproceedings AHL+26

---

CHI 2026

ACM CHI Conference on Human Factors in Computing Systems. Barcelona, Spain, Apr 13-17, 2026.

---

Authors

L. B. Andersen • M. Herklotz • A. Liu • M. Goeke • M. Juelich • C. Kern • F. Kreuter

---

Links

DOI

---

Research Area

 C4 | Computational Social Sciences