RESEARCH SOURCES

Valuable Sources for AI Governance

Comprehensive collection of authoritative resources, research findings, and best practices from leading international organisations, governments, and academic institutions.

Evidence-based foundations for responsible AI implementation across all sectors.

International organisations

🌐

UNESCO

United Nations Educational, Scientific and Cultural organisation

AI Ethics and Governance Lab

Knowledge hub bringing together case studies, good practices, and cutting-edge research. Focus on framing key issues, providing tangible insights from practice, introducing innovative tools, and policy recommendations.

Visit Resource

Guidance for Generative AI in Education and Research

First global guidance on GenAI in education (April 2025). Supports countries in implementing immediate actions, planning long-term policies, and developing human capacity.

Visit Resource

AI Competency Frameworks

Two comprehensive frameworks:

  • AI Competency Framework for Students
  • AI Competency Framework for Teachers

Purpose: Guide countries in supporting students and teachers to understand AI potential, develop AI literacy, and enable safe and ethical AI integration in education.

🏥

World Health organisation (WHO)

Leading authority on global health

Ethics and Governance of Artificial Intelligence for Health (June 2021)

150-page comprehensive guidance document developed over 18 months with leading experts in ethics, digital technology, law, human rights, and Ministries of Health.

Six Consensus Principles for AI in Health:

  • Ethics and human rights at the heart of design, deployment, and use
  • Public benefit for all countries
  • Accountability of stakeholders (public and private sector)
  • Responsiveness to healthcare workers
  • Community and individual health protection
  • Governance maximising promise while managing risks
Visit Resource
📊

OECD

Organisation for Economic Co-operation and Development

OECD AI Principles (2019)

First intergovernmental standard on AI, promoting AI that is innovative and trustworthy and that respects human rights and democratic values.

Visit Resource
🌍

United Nations

Global governance and sustainable development

Harnessing AI for the Sustainable Development Goals (SDGs)

Framework for leveraging AI to accelerate achievement of SDGs across health, education, energy, climate action, and biodiversity.

Visit Resource

UN Global Issues: Artificial Intelligence

Comprehensive overview of AI's role in addressing global challenges and the need for ethical frameworks.

Visit Resource

Financial & Economic Institutions

💰

International Monetary Fund (IMF)

AI Projects in Financial Supervisory Authorities (October 2025)

Working Paper No. 2025/199 - 34 pages examining how financial supervisory authorities enhance their toolkit through AI adoption.

Key Challenges Identified:

  • Ensuring explainability of AI decisions
  • Mitigating algorithmic bias in financial decisions
  • Stakeholder collaboration requirements
  • Robust governance frameworks necessity
  • Adequate resource allocation
Visit Resource
🤝

World Economic Forum (WEF)

IBM Responsible AI Case Study (September 2021)

Co-authored with Markkula Centre for Applied Ethics at Santa Clara University. Examines IBM's comprehensive approach to ethical AI.

Five Open-Source Toolkits:

  • AI Explainability 360: 8 algorithms for making ML models more explainable
  • AI Fairness 360: 70 fairness metrics + 10 bias-mitigation algorithms
  • Adversarial Robustness Toolbox: Tools for overcoming adversarial attacks
  • AI FactSheets 360: Transparency documentation methodology
  • Uncertainty Quantification 360: Tools to test reliability of AI predictions
Visit Resource

Advancing Responsible AI Innovation: A Playbook (2025)

Comprehensive guide for organisations implementing responsible AI practices.

Visit Resource

Defence & National Security

🛡️

NATO

North Atlantic Treaty organisation

Summary of NATO's Revised AI Strategy (July 2024)

First adopted in October 2021, revised in July 2024. Includes six Principles of Responsible Use (PRUs) for AI in Defence.

Six Principles of Responsible Use:

  • Lawfulness: AI applications developed and used in accordance with national and international law
  • Responsibility and Accountability: Clear human responsibility for AI systems
  • Explainability and Traceability: AI decisions can be understood and traced
  • Reliability: AI systems perform as intended
  • Governability: AI systems can be controlled and managed
  • Bias Mitigation: Active measures to prevent and address bias
Visit Resource
🇺🇸

U.S. Department of Defence

Responsible AI (2020)

Five DoD AI Ethical Principles: Responsible, Equitable, Traceable, Reliable, and Governable.

Visit Resource

U.S. Intelligence Community AI Ethics Principles

Institutional ethical, legal, and accountability frameworks. Consistent and enforceable ethical frameworks for national security AI use.

Visit Resource

Government AI Implementation

🇪🇪

Estonia AI Implementation Case Study

Government leverages AI to enhance public services, streamline operations, and improve citizen engagement through a digital-first government model.

Key Features:

  • 50+ AI-powered solutions integrated into public services
  • Human-in-the-loop approach ensures accountability
  • Collaboration with private sector, academia, and civil society
  • GDPR and fundamental rights compliance
Visit Resource

Media & Creative Industries

🎬

Deepfake Governance

Global Governance Challenges of Deepfake Technology

Comprehensive analysis of ethical concerns, legal accountability challenges, and governance frameworks for machine-manipulated media.

Visit Resource

AI Impact on Creative Industries (February 2024)

WEF report on how AI will augment existing creative jobs, create new fields and roles, and lower barriers to entry for creative work.

Visit Resource

Common Governance Principles Across All Sectors

🛡️

Accountable Governance and Ethical Leadership

Establish clear accountability structures and ethical oversight for AI systems.

🧭

Purpose-Driven and Rights-Based Innovation

Ensure technological advancement is intentional, beneficial, and respects fundamental human rights.

👤

Human Agency and Oversight

Augment human capabilities while preserving dignity and meaningful control.

🔒

Systemic Safety and Reliability

Engineer systems to be safe, secure, and reliable across diverse and adversarial conditions.

💾

Data Integrity and Privacy

Ensure training data is accurate, representative, and handled with strict privacy protocols.

🔍

Algorithmic Transparency and Explainability

Disclose system capabilities and provide comprehensible rationale for automated decisions.

📊

Continuous Assurance and Adaptation

Implement ongoing monitoring to ensure systems remain safe, effective, and ethically aligned.

⚖️

Equitable Impact and Societal Well-being

Assess and govern the systemic effects of AI on social equity and collective psychological health.

🤝

Ecosystem Accountability and Market Fairness

Promote a healthy, fair, and diverse technological ecosystem free from anti-competitive practices.

🌐

Global and Cultural Context

Ensure governance architectures are culturally aware, adaptable, and globally inclusive.

Verified Footnotes Index

To maintain absolute transparency, all core statistics, claims, and compliance references used across the ETHOS Institute platform are systematically mapped here to their peer-reviewed publications, legal statutes, and empirical surveys.

[1]78% AI Adoption: McKinsey Global Institute, "The State of AI in 2025: Generative AI Adoption Waves" (reporting that 78% of surveyed organisations have integrated AI capabilities into at least one business function).
[2]15–40% Effective Governance Gap: ISO/IEC 42001:2023 / NIST AI RMF 1.0 Assessment Surveys (revealing that while 78% of firms adopt AI, only 15% to 40% have active, board-governed risk-management safeguards).
[3]2x Profit Premium: McKinsey & Company, "How AI Leaders Outperform Peers" (companies implementing trustworthy AI governance frameworks achieve double the profit premium of lagging peers).
[4]28% Fewer Failures: NIST AI Risk Management Framework 1.0 Impact Reports (organisations applying systematic bias testing, documentation, and continuous verification controls experience 28% fewer automated system failures).
[5]69% Consumer Distrust: Edelman Trust Barometer (reporting that 69% of global consumers express distrust in autonomous AI systems lacking visible, independent ethical certifications).
[6]€35M or 7% Non-compliance Fines: Official EU AI Act Text, Article 99 (detailing penalties for non-compliance with prohibited AI practices or governance safeguards of up to €35 million or 7% of global annual turnover, whichever is higher).
[7]30+ Years of Research: Grounded in peer-reviewed scientific monographs, specifically "AI Ethics: A Historical-Comparative Approach" (Edward Elgar Publishing, 2026), providing rigorous historical and theoretical validation.

Explore Real-World Applications

See how these principles are being implemented across industries and sectors worldwide.

View Use CasesExplore the Framework