580 California St., Suite 400
San Francisco, CA, 94104
Key research themes
1. How can ethical frameworks and governance shape the development of AI for social good?
This research area investigates the formulation, application, and impact of ethical principles and governance mechanisms aimed at guiding AI technologies towards beneficial societal outcomes. It addresses the translation of abstract ethical guidelines into concrete policy recommendations, the role of multi-stakeholder collaboration involving governments, industry, and academia, and the challenges of embedding values such as human dignity, fairness, and transparency into AI systems. This theme matters because the widespread societal adoption of AI raises complex ethical risks and opportunities that require proactive, principled, and coordinated governance to realize AI’s potential for social good while mitigating harms.
Key finding: This study identifies four fundamental opportunities AI offers society—promoting human autonomy, agency, capabilities, and societal cohesion—and pairs these with corresponding risks, formulating five ethical principles... Read more
Key finding: Through comparative analysis of national AI strategy reports from the US, EU, and UK, this paper reveals that while these documents address various ethical, social, and economic aspects, they lack an overarching political... Read more
Key finding: This work distinguishes two critical approaches to promoting beneficial AI—extrinsic measures (external incentives or constraints) and intrinsic measures (motivation within AI researchers)—arguing that intrinsic factors,... Read more
Key finding: Introducing the concept of harmony rooted in East Asian philosophical traditions as a complementary ethical principle for AI, this study argues that incorporating harmony—entailing well-balanced relationships, situational... Read more
2. What are effective models and methodologies for integrating AI into social good initiatives and domain-specific applications?
This theme encompasses the technical, organizational, and interdisciplinary methodologies for applying AI to concrete social challenges, including education, healthcare, environmental sustainability, and rural empowerment. It covers frameworks for collaboration between AI experts and domain specialists, socially inclusive design strategies, and the development of evidence-based AI tools that respond sensitively to user needs and sociocultural contexts. This theme addresses the implementation challenges and best practices that maximize AI’s positive impact on targeted populations and social sectors.
Key finding: This paper synthesizes guidelines for establishing sustainable interdisciplinary partnerships between AI researchers and domain experts, emphasizing early domain involvement and ethical adherence. It reviews diverse AI4SG... Read more
Key finding: Proposes an innovative evidence-informed AI framework blending clinical psychology models (CBT, DBT, CFT, ACT, narrative therapy) with real-world user data and emotional tone calibration to build AI companions that offer... Read more
Key finding: Presents a practical AI-powered web application using NLP and ML to provide personalized educational and scholarship opportunities for underserved rural students, addressing digital divide challenges. Developed with... Read more
Key finding: This educational resource details a structured, multi-institutional course designed to engage diverse learner groups in data science for social good, combining theoretical foundations with case study analyses and ethical... Read more
3. How can participatory, community-centric, and contestation-based approaches democratize AI governance to better serve social good?
This research theme explores mechanisms that empower communities and stakeholders traditionally marginalized by centralized AI development and governance. It investigates social choice dilemmas in aggregating diverse societal ethical views, the limitations of technocratic ethics approaches, and proposes tactical frameworks for civic engagement, activism, and collective AI accountability. It also examines culturally grounded worldviews and participatory design paradigms as means to flatten hierarchical knowledge and power structures in AI development. This theme matters because democratizing AI governance opens avenues for legitimacy, justice, and responsiveness in AI’s societal impact.
Key finding: Analyzes the foundational social choice challenges facing AI ethics based on aggregating societal ethical views, including standing (whose views count), measurement (how views are identified), and aggregation (how they are... Read more
Key finding: This paper introduces the concept of algorithmic contestation, proposing ten 'Rules for Radical AI' inspired by Saul Alinsky’s organizing tactics to empower activists and communities to challenge AI power structures. Case... Read more
Key finding: Critically examines prevailing human-centered AI narratives that often perpetuate Western dualist philosophical legacies, arguing for inclusion of stakeholder- and community-centric value frameworks such as Ubuntu and Maximum... Read more