Greening AI Inference with Accuracy and Latency-aware User Incentives

2026-05-26

AI Summary

This work addresses the high carbon emissions of AI inference services and the lack of incentive mechanisms in existing systems that jointly account for user preferences regarding accuracy, latency, and environmental impact. The authors propose a carbon-aware AI inference framework that, for the first time, integrates users’ trade-offs between service quality (accuracy and latency) and environmental consciousness into scheduling design. By introducing a two-tier subscription pricing model, the framework incentivizes users to voluntarily accept reduced service quality during high-carbon-intensity periods in exchange for discounts. Combining carbon-intensity-aware scheduling, tiered service quality levels, and user preference modeling, the approach flexibly adapts to diverse models and resource conditions, significantly reducing carbon emissions while preserving user experience and offering a practical, commercially viable pathway toward greener AI services.

Abstract

The widespread use of AI services has raised concerns for its environmental sustainability, towards which recent studies have identified carbon emissions of AI inference as the major contributor. This paper introduces a framework for designing AI inference incentives based on the users' valuation for inference quality and latency, together with their environmental consciousness, while accounting for the tradeoff between carbon emissions and the two QoE parameters. Our approach can accommodate different tradeoffs, that depend on the size and complexity of the AI models and the allocation of resources to serve inference requests. The incentives can be offered through a practical two-tier service subscription that offers users a discount in exchange for reduced carbon emissions. The discounted service option gives the AI provider the flexibility to serve some percentage of inference requests at a lower quality and higher latency during periods of high carbon intensity.

Problem

• Green AI
• Carbon Emissions
• AI Inference
• User Incentives
• Environmental Sustainability

Innovation

• Green AI
• Inference Optimization
• User Incentives
• Carbon-aware Computing
• Quality of Experience (QoE)