pro social ai

By the Prosocial AI editorial team.

Executive Summary

This literature review explores the evolving field of prosocial AI, highlighting its potential to align technological advancements with societal well-being and ethical behavior. Prosocial AI prioritizes human values, fostering cooperation, equity, and ethical decision-making to address pressing social challenges such as healthcare accessibility, education inequality, and environmental sustainability. This paper examines various academic frameworks and models that shape prosocial AI development, focusing on mitigating risks, promoting inclusivity, and serving as a catalyst for global well-being. Insights from industry discussions and global forums, such as the United Nations General Assembly, underscore AI’s potential to transform health systems and promote holistic well-being.

 Literature Review

 Human-Machine Interaction and Prosocial Behavior

Nielsen et al. (2022) explored how artificial intelligence impacts human-machine interaction through the Computers Are Social Actors (CASA) framework. According to CASA, humans unconsciously apply social norms to machines, treating them as social entities. Nielsen et al. found that human-like AI can evoke prosocial behaviors such as empathy, cooperation, and trust. Nielsen et al.’s findings suggest that AI systems designed to model human social dynamics can foster ethical and cooperative behaviors in users, particularly in sectors like education, healthcare, and customer service (Nielsen et al., 2022).

 AI in Child Development

Peter et al. (2021) investigated how social robots influence children’s prosocial behavior, grounding their study in Social Cognitive Theory (Bandura, 1986), which posits that behaviors are learned through observation. In their study, children aged 8 to 10 interacted with robots, displaying prosocial behaviors such as sharing and cooperation. The study found that children exposed to these robots increased their prosocial behaviors. These findings indicate that AI can act as a role model, fostering cooperation and social learning, particularly in educational contexts where ethical behavior can be modeled and promoted (Peter et al., 2021).

 Collective Agency and AI

Bryson (2015) examined AI’s role in promoting collective agency, a social and political philosophy concept that refers to the capacity of groups to act together toward shared goals. Bryson argued that AI, as an extension of human agency, can facilitate collective actions that benefit society. AI systems designed to model altruistic and cooperative behaviors can contribute to collective well-being, offering strategies to improve social cohesion, address global challenges, and foster ethical decision-making on individual and collective levels (Bryson, 2015).

 Prosocial Rule Breaking and Ethical AI

Ramanayake et al. (2022) introduced the concept of Prosocial Rule Breaking (PSRB), illustrating how AI can be programmed to break the rules ethically when doing so benefits society. For instance, in healthcare, professionals often deviate from protocols to save lives or improve patient outcomes. Similarly, AI systems could break preset rules in high-stakes environments, such as autonomous vehicles or healthcare, when the situation warrants it. This framework promotes ethical flexibility and contextual awareness, allowing AI to prioritize societal well-being over rigid rule-following (Ramanayake et al., 2022).

 AI in Global Health and Holistic Well-Being

At the 79th United Nations General Assembly (UNGA), the Center for Integrative Global Oral Health (CIGOH) experts emphasized how prosocial AI can catalyze holistic health. They advocated a transition from disease-centric healthcare to a salutogenic model, focusing on factors supporting overall well-being. When framed as a social determinant of health, AI can expand healthcare access in underserved areas and empower individuals to manage their health. AI-driven tools in mental health and personalized diagnostics demonstrate how AI can revolutionize healthcare delivery and reduce health disparities (CIGOH at the UNGA79, 2024).

 The Four T’s of Prosocial AI: Tailored, Trained, Tested, and Targeted

Walther (2024) outlined the “Four T’s” framework—Tailored, Trained, Tested, and Targeted—as essential for developing prosocial AI systems that are ethical and inclusive. Tailored AI solutions address specific societal needs, such as healthcare in rural areas. Trained AI relies on diverse datasets to prevent bias, ensuring inclusivity. Rigorous testing, including ethical audits and stress tests, ensures AI systems operate fairly and reflect societal values. Finally, targeted AI focuses on measurable outcomes like reducing environmental impact or increasing education access, ensuring that AI serves the public good while mitigating risks (Walther, 2024).

 The PRO Framework: Purpose-Driven Growth, Resilience, and Optimization

The PRO Framework, outlined by GetCo AI (2024), offers a roadmap for businesses to integrate prosocial AI into their operations. It emphasizes Purpose-driven growth, Resilience and risk mitigation, and Optimization for inclusive impact. By aligning AI with broader societal goals, companies can contribute to social good while driving business success. AI tools in mental health, healthcare diagnostics, and recruitment exemplify how prosocial AI can bridge critical healthcare access gaps and promote hiring fairness. The PRO Framework advocates collaboration between businesses, NGOs, and academic institutions to ensure AI systems are inclusive and socially responsible (GetCo AI, 2024).

Discussion

 Balancing Innovation and Ethical Responsibility

A central theme in prosocial AI is the balance between technological innovation and ethical responsibility. AI systems must be designed to avoid reinforcing biases or exacerbating inequalities as they become more integrated into everyday life. Walther (2024) stressed the importance of using diverse datasets and conducting rigorous testing to prevent bias and ensure AI systems reflect societal values. The ethical frameworks, like the “Four T’s,” guide developers in addressing these issues. However, ethical AI development requires ongoing vigilance, collaboration, and transparency among technologists, ethicists, and policymakers (Walther, 2024).

Nielsen et al. (2022) demonstrated how human-like AI can evoke prosocial behaviors, but this also raises ethical concerns. The potential for AI to manipulate human behavior must be addressed by ensuring transparency and prioritizing users’ well-being. Developers must carefully navigate these ethical dilemmas, especially when using AI in sensitive healthcare or education areas where user trust and autonomy are paramount (Nielsen et al., 2022).

 AI as a Catalyst for Holistic Health

Prosocial AI’s integration into healthcare represents a transformative opportunity for promoting well-being. At the UNGA79, experts highlighted AI’s potential to improve healthcare access, especially in underserved areas. AI tools such as mental health platforms and personalized diagnostics demonstrate how AI can reduce health disparities and empower individuals to manage their health. However, integrating AI into healthcare also raises ethical challenges regarding accountability and transparency. Systems that make life-and-death decisions must undergo rigorous ethical scrutiny to prevent harm and ensure equitable healthcare delivery (CIGOH at the UNGA79, 2024).

Ramanayake et al. (2022) proposed Prosocial Rule Breaking (PSRB) as a framework to allow AI systems the ethical flexibility to break the rules when necessary for the greater good. This framework is critical in healthcare environments where strict rule adherence may not always yield the best outcomes, emphasizing the need for adaptable AI to respond to complex ethical dilemmas (Ramanayake et al., 2022).

 The Role of Business in Advancing Prosocial AI

Businesses play a crucial role in advancing prosocial AI. The PRO Framework encourages companies to align their AI systems with societal goals while maintaining profitability. However, businesses must ensure that their prosocial AI efforts are not merely marketing strategies but are backed by real investment in ethical development. Walther (2024) argued that the future of AI lies in its ability to foster innovation, sustainability, equity, and well-being. Success in prosocial AI requires businesses to shift their focus from purely financial metrics to a broader definition of value that includes social impact.

 Future Directions for Prosocial AI

The future of prosocial AI holds both opportunities and challenges. As industries increasingly adopt prosocial AI frameworks, the technology has the potential to address global issues such as climate change, healthcare accessibility, and education inequality. However, achieving lasting change will require sustained commitment from businesses, governments, and civil society. Cross-sector collaboration ensures that AI systems are developed and deployed ethically. Additionally, the continuous refinement of ethical frameworks and mechanisms for accountability will be critical to the long-term success of prosocial AI. Clear guidelines on data privacy, ethical decision-making, and transparency will help embed prosocial principles at the core of AI development.

 References

Bryson, J. J. (2015). Artificial intelligence and prosocial behavior. In C. Misselhorn (Ed.), Collective agency and cooperation in natural and artificial systems: Explanation, implementation, and simulation (pp. 281–306). Springer. https://doi.org/10.1007/978-3-319-15515-9_15

CIGOH at the UNGA79. (2024, September 23). Prosocial AI as a catalyst of holistic health for all. Center for Integrative Global Oral Health. https://www.dental.upenn.edu/news-events/2024/09/23/cigoh-at-the-unga79-prosocial-ai-as-a-catalyst-of-holistic-health-for-all/

GetCo AI. (2024). Why are more leaders emphasizing prosocial AI to guide product development? GetCo AI. https://getcoai.com/news/why-more-leaders-are-emphasizing-prosocial-ai-to-guide-product-development/

Nielsen, Y. A., Pfattheicher, S., & Keijsers, M. (2022). Prosocial behavior toward machines. Current Opinion in Psychology, 43, 260–265. https://doi.org/10.1016/j.copsyc.2021.08.004

Peter, J., Kühne, R., & Barco, A. (2021). Can social robots affect children’s prosocial behavior? Computers in Human Behavior, 120, 106712. https://doi.org/10.1016/j.chb.2021.106712

Ramanayake, R., Wicke, P., & Nallur, V. (2022). Immune moral models? Prosocial rule breaking as a moral enhancement approach for ethical AI. AI & Society, 38(3), 801–813. https://doi.org/10.1007/s00146-022-01478-z

Walther, C. C. (2024). How to reduce the risk of AI and enhance its societal impact. Knowledge@Wharton. https://knowledge.wharton.upenn.edu/article/how-to-reduce-the-risk-of-ai-and-enhance-its-societal-impact/