HBKU - FRG Grant

New Grant Award - Islamic Media Ethics in the Digital Age

Islamic Media Ethics in the Digital Age

AI-Powered Solutions for Combating Misinformation and Promoting Constructive Cross-Cultural Dialogue

This project, supported by the HBKU Flagship Research Grant (3rd Cycle), investigates how Islamic media ethics can be systematically integrated into AI-driven language and multimodal systems to address misinformation, polarization, and ethical failures in today’s digital information environment.

The project runs from 2025–2027 and brings together expertise from AI and Arabic NLP, Islamic studies, media ethics, and public policy.

Project motivation

Current AI systems for misinformation detection and content moderation are often developed using generic or culturally narrow assumptions, limiting their effectiveness and trustworthiness in diverse societies. In particular, Islamic ethical principles such as truthfulness (ṣidq), justice (ʿadl), responsibility (amānah), and avoidance of harm in Islamic sense are rarely operationalized in computational systems.

This project aims to close this gap by designing ethically grounded, culturally aware AI systems that reflect Islamic media ethics while remaining compatible with global norms such as human rights and international law.

Research objectives

The project pursues four core objectives:

  1. Operationalize Islamic media ethics for AI
    Translate ethical principles from Islamic scholarship into computationally testable criteria for AI systems.

  2. Develop culturally grounded benchmarks and datasets
    Create evaluation resources for misinformation, harmful content, and ethical reasoning grounded in Islamic and cross-cultural contexts.

  3. Build AI models with ethical awareness
    Design language and multimodal models capable of context-aware fact-checking, responsible moderation, and calibrated uncertainty.

  4. Support constructive digital dialogue
    Enable AI-assisted platforms that promote respectful, informed, and dialogical engagement across cultures.

Methodology

The project adopts an interdisciplinary and multi-layered approach:

  • Source grounding: Islamic texts, trusted scholarly sources, international law, and human rights frameworks
  • AI & NLP techniques: multilingual and multimodal LLMs, Arabic NLP, explainability, and automated reasoning
  • Evaluation: benchmark-driven assessment, human-in-the-loop validation, and ethics-focused metrics

Expected outcomes

Key outcomes include:

  • Public datasets and benchmarks for ethical AI and misinformation research
  • Evaluation frameworks for culturally and ethically grounded LLMs
  • AI models and tools for fact-checking, content moderation, and dialogue facilitation
  • Digital literacy resources aligned with Islamic principles of critical thinking

All releases will follow open and responsible research practices.

Team and collaboration

The project is led by researchers from HBKU, QCRI and Doha Institute for Graduate Studies, with contributions from:

  • Artificial Intelligence and Arabic NLP
  • Islamic studies and media ethics
  • Public policy and cross-cultural communication

Resources and updates

Datasets, models, and research outputs will be released progressively through the following collection:

🔗 Islamic Knowledge in LLMs
Islamic Knowledge in LLMs

Updates on milestones, publications, and releases will be posted on this page throughout the project.

References