Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

FairSense-AI: Responsible AI Meets Sustainability

Shaina Raza, Mukund Sayeeganesh Chettiar, Matin Yousefabadi, Tahniat Khan, Marcelo Lotif

TL;DR

FairSense-AI addresses the dual challenge of societal bias and the environmental footprint of large-scale AI by delivering a multimodal framework that detects and mitigates bias in both text and images. Leveraging LLMs and VLMs, it provides bias scores, explanations, and automated recommendations, alongside an AI risk assessment component aligned with MIT AI Risk Repository and NIST AI RMF. The system emphasizes Green AI through pruning and mixed-precision techniques, with energy-use monitoring to reduce emissions. Through practical case studies, it demonstrates how content fairness and sustainability can be integrated into real-world AI workflows, offering an API and tooling for scalable bias auditing and governance.

Abstract

In this paper, we introduce FairSense-AI: a multimodal framework designed to detect and mitigate bias in both text and images. By leveraging Large Language Models (LLMs) and Vision-Language Models (VLMs), FairSense-AI uncovers subtle forms of prejudice or stereotyping that can appear in content, providing users with bias scores, explanatory highlights, and automated recommendations for fairness enhancements. In addition, FairSense-AI integrates an AI risk assessment component that aligns with frameworks like the MIT AI Risk Repository and NIST AI Risk Management Framework, enabling structured identification of ethical and safety concerns. The platform is optimized for energy efficiency via techniques such as model pruning and mixed-precision computation, thereby reducing its environmental footprint. Through a series of case studies and applications, we demonstrate how FairSense-AI promotes responsible AI use by addressing both the social dimension of fairness and the pressing need for sustainability in large-scale AI deployments. https://vectorinstitute.github.io/FairSense-AI, https://pypi.org/project/fair-sense-ai/ (Sustainability , Responsible AI , Large Language Models , Vision Language Models , Ethical AI , Green AI)

FairSense-AI: Responsible AI Meets Sustainability

TL;DR

FairSense-AI addresses the dual challenge of societal bias and the environmental footprint of large-scale AI by delivering a multimodal framework that detects and mitigates bias in both text and images. Leveraging LLMs and VLMs, it provides bias scores, explanations, and automated recommendations, alongside an AI risk assessment component aligned with MIT AI Risk Repository and NIST AI RMF. The system emphasizes Green AI through pruning and mixed-precision techniques, with energy-use monitoring to reduce emissions. Through practical case studies, it demonstrates how content fairness and sustainability can be integrated into real-world AI workflows, offering an API and tooling for scalable bias auditing and governance.

Abstract

In this paper, we introduce FairSense-AI: a multimodal framework designed to detect and mitigate bias in both text and images. By leveraging Large Language Models (LLMs) and Vision-Language Models (VLMs), FairSense-AI uncovers subtle forms of prejudice or stereotyping that can appear in content, providing users with bias scores, explanatory highlights, and automated recommendations for fairness enhancements. In addition, FairSense-AI integrates an AI risk assessment component that aligns with frameworks like the MIT AI Risk Repository and NIST AI Risk Management Framework, enabling structured identification of ethical and safety concerns. The platform is optimized for energy efficiency via techniques such as model pruning and mixed-precision computation, thereby reducing its environmental footprint. Through a series of case studies and applications, we demonstrate how FairSense-AI promotes responsible AI use by addressing both the social dimension of fairness and the pressing need for sustainability in large-scale AI deployments. https://vectorinstitute.github.io/FairSense-AI, https://pypi.org/project/fair-sense-ai/ (Sustainability , Responsible AI , Large Language Models , Vision Language Models , Ethical AI , Green AI)

Paper Structure

This paper contains 13 sections, 1 equation, 4 figures, 4 algorithms.

Table of Contents

  1. Introduction
  2. Theoretical and Scientific Foundations
  3. Computational Linguistics and Social Psychology
  4. Multimodal Learning and Vision-Language Integration
  5. Environmental Sustainability and "Green AI"
  6. Technical Approach
  7. FairSense-AI Methodology
  8. FairSense-AI Pipeline
  9. Algorithm
  10. Implementation Details fair-sense-ai library and use its API to analyze data programmatically. The API includes functions
  11. Applications and Case Studies
  12. Comparative Analysis
  13. Conclusion and Future Works

Figures (4)

  • Figure 1: FairSense-AI platform
  • Figure 2: Text analysis
  • Figure 3: Image analysis
  • Figure 4: The Risk Identification and Mitigation interface in FairSenseAI. Users can input an AI project description in the text field (top) and then click Analyze Risks to retrieve potential AI risks from the MIT repository and relevant mitigation strategies from the NIST AI RMF. The identified risks from the MIT repository are displayed below, while the complete results, including corresponding NIST mitigation strategies, are exported as a downloadable CSV.