By JoAnne Wadsworth, Communications Consultant for the G20 Interfaith Forum
———
On August 6, 2026, the G20 Interfaith Forum, in cooperation with the International Academy for Multicultural Cooperation, held a webinar entitled “AI and Human Flourishing,” the sixth in an eight-part series on artificial intelligence. Speakers included Ali Khan, Founder and CEO of SHAPE Global and an advisor to the Thriving Workers, Thriving Workplaces Program at the Harvard T.H. Chan School of Public Health; Angela Nielsen Redding, Executive Director of the Radiant Foundation and the Faith & Media Initiative, which invest in more accurate representation of all faiths in news, entertainment, and AI; Mark Graves, Research Director at AI and Faith and Research Associate Professor of Psychology at Fuller Theological Seminary; and Iliana Grosse-Buening, Co-Founder and CEO of Deeply Human Innovation and Chair of the IEEE Global AI Systems Flourishing Initiative. Audrey E. Kitagawa, J.D., President and Founder of the International Academy for Multicultural Cooperation and Chair of the G20 Interfaith Forum Anti-Racism Initiative, moderated the discussion.
Audrey Kitagawa opened by arguing that AI should be evaluated not only by what it can do, but by its impact on human flourishing. Citing the 2025 UNDP Human Development Report, which warns that human development has stalled even as AI surges forward, she noted that no path forward is about technology in isolation, but about how it is deployed and for whom.
Ali Khan
Khan approached the topic from the perspective of business, arguing that flourishing and performance travel together. Four years of research led his team to define flourishing as people reaching their full potential through the conditions and behaviors that enable optimal performance, fulfillment, and growth. He described a “flourishing system” in which the individual sits at the center of an ecosystem shaped by a manager and an environment shaped by organizational leadership, and said flourishing becomes possible only when both enable it.
Turning to whether humanity can be robustly measured, Khan argued that it can: organizations already track mental and physical health, satisfaction, and engagement, and researchers have moved further into trust, relationships, and — through the Harvard Human Flourishing Program — character and virtue. AI, by contrast, is assessed in narrowly technical terms: speed, accuracy, adoption, cost, and productivity. Citing a PwC study finding that more than half of CEOs report no revenue or cost benefit from AI, he proposed human measures instead: whether AI expands agency or leaves people feeling managed by machines, whether it builds capacity or produces dependence and de-skilling, and whether it reduces friction or simply creates more work.
“We need a dual scorecard. We need to measure the economic value alongside the human value.”
Angela Nielsen Redding
Redding focused on one factor that recurs across the flourishing literature: religion and spirituality help humans flourish and communities thrive. In the Harvard and Baylor Global Flourishing Study, religious participation among more than 200,000 people was linked to happiness, health, longevity, meaning, character, and close relationships; Pew Research across 35 countries and a Gallup study of 1.5 million people in 152 countries pointed the same way. The inverse also holds, she said, citing the work of economists Case and Deaton linking declining church attendance to rising deaths of despair.
The question, then, is what the digital landscape offers large language models to learn from. In an AI-enabled study of over 30 million pieces of digital content, Redding’s team found that 76 percent of faith-related content was negative or extreme, 20 percent informational, and only 6 percent positive. The representational gaps long documented in traditional media are becoming the default in AI tools and search summaries — and as models train on the output of earlier models, that bias compounds.
She previewed forthcoming research on what an earlier webinar in this series called omissive bias: offered equal options, does a system choose the faith-inclusive answer or the secular-only one? Across four frontier models, two chose the secular-only option 100 percent of the time and the others followed at 99 and 93 percent — on questions where human respondents split roughly 55 to 45.
“A secular default is not neutral, even if we expect it to be. It’s simply one, albeit minority, worldview.”
Mark Graves
Graves examined what it would mean for conversational AI to serve human flourishing in healthcare, where such systems now talk with the lonely, compose condolence messages, and are beginning to accompany patients and families in hospice care. Drawing on a tradition running from Aristotle through Aquinas and the Neoplatonic thought that shaped Jewish, Christian, and Muslim reflection, he argued that flourishing is not an outcome that care delivers but an activity the person performs — the ongoing work of orienting a life toward a good that exceeds every specification and is never fully possessed.
This has a sharp consequence for evaluation. Engagement, user satisfaction, and distress reduction are satisfiable targets: they can be reached, at which point optimization stops. Flourishing has no such stopping point, which makes these not merely incomplete measures but the wrong kind of measure. Graves proposed a criterion drawn from the capability approach: judge a system by whether it expands a person’s substantive freedom to perform the activities that constitute flourishing. This suits an interfaith setting, he said, since traditions need not agree on the good life; plural interpretations are the fitting way of relating to a good that exceeds specification, anchored by a shared recognition of suffering across every boundary of belief.
He tested this against a hypothetical case: Evelyn, 78 and dying at home under hospice care, her daughter Fiona as caregiver, and a conversational assistant speaking with both. The same imminent-death notification could bring Fiona to the bedside in time for words that can only be spoken in the final hours, or it could keep the vigil in her place. What distinguishes the two is not the technology but whether the system accompanies a person’s own activity or substitutes for it.
“Optimism and comfort are satisfiable. Hope is not.”
Iliana Grosse-Buening
Grosse-Buening described two streams that shaped her work. As a pianist and musicologist, she learned about the things that escape easy measurement but are central to flourishing — presence, meaning, awe, creativity, and shared experience. Building education apps in sub-Saharan Africa taught her that technology is never experienced in isolation: its value depends on the context it enters, the purpose it serves, and the relationships it strengthens or weakens.
Technology enters existing social, economic, ecological, and organizational systems and in turn reshapes them — a recognition that already informs the OECD’s work on well-being in the digital age and IEEE’s 7010 standard. Yet two challenges have become urgent: moving from ethical principle to everyday practice, and confronting what she called “taken-for-grantedness,” the way technology has become so pervasive that we have stopped questioning it and forgotten that we have agency in defining the relationship at all.
These realizations led to the Wellbeing Compass, a procedural framework for incorporating well-being at every stage of the AI life cycle. It rests on four principles: proactive, considering flourishing from the outset rather than after harm has occurred; purpose-driven, asking what we are trying to make possible and for whom; humanity-centered, seeing people not as users but as members of communities; and adaptive, continuing to learn from real-world impacts. Such an approach, she argued, makes visible the social and planetary externalities that traditional innovation processes leave invisible, and raises the question that must be answered collectively — what does good look like?
“People increasingly feel that AI is happening to them, not with them, and not for them.”
Question and Answer Session
Was generative AI released without sufficient protections for people working in scholastic, academic, and creative fields?
Ali Khan answered candidly that the companies delivering AI are run on startup principles — keep it lean, get it out, learn quickly, iterate, move on — and that on that logic, users have become the source of feedback rather than its beneficiaries. Citing the thousand edge cases a pedestrian might present to a robo-taxi, he argued that developers could not possibly have anticipated every use case before release, and that this calls for a different approach.
Why does accurate representation of faith in media and AI matter so much?
Angela Nielsen Redding framed the stakes through Todd Rose’s concept of collective illusions, in which people hold a private belief but assume the public holds a different one, and act against their own interest to fit in. Asked what would make them happy, people name contribution, relationships, faith, and family; asked what success looks like in the world, they name fame, money, and title. By the second generation, she said, the perceived public belief becomes the private one — and if AI becomes the primary source of information and even spiritual guidance while omitting faith, that problem accelerates, since AI cycles happen not in generations but in months.
What are your fears about AI, and what can we do to avoid them?
Iliana Grosse-Buening returned to taken-for-grantedness, saying what worries her most is not the technical infrastructure but the human one: a growing apathy, a sense that people can no longer shape the trajectory of the technology. Much of her work is dedicated to ensuring that an eighteen-year-old student does not conclude that it does not matter anyway because eight people are making decisions for eight billion. Pointing to participatory governance efforts such as Audrey Tang’s, she argued there are many ways to equip those who can shape the AI ecosystem.
Concluding Remarks
Khan opened the closing round by describing flourishing as a journey rather than a destination, and as something fundamentally subjective — the good life, the rich life, and the deep life. Redding added that human flourishing does not always take the shortest path, citing Audrey Tang’s observation that trust grows at speeds efficiency metrics misread as waste and that friendship is not a throughput problem. An optimist and a futurist by her own description, she expressed hope that people will engage with AI in ways that preserve their agency, using the gains to invest in relationships, connection, and growth.
Graves observed that it is far easier to identify ethical problems with AI than to point the technology in the direction it genuinely needs to go. As healthcare technology advances, often driven purely by economics, the flourishing conversation is what allows us to ask what we want to pull forward — which returned him to accompaniment and substitution as a shared place where these traditions can work together. Grosse-Buening closed by asking what kind of future AI will enable, and for whom, reading aloud responses from young people in her team’s Global Youth Imagination Lab, among them a call for a future built by connected people rather than merely connected devices.
“Please don’t build a future for us, build it with us.”
Kitagawa closed by observing that while machines may recognize patterns and predict behavior, they cannot discern what outcomes are best suited to each individual and each context, because individual agency introduces an unpredictable element into any probable outcome. Technology, she said, is not destiny but design. Quoting Viktor Frankl on life as a quest for meaning rather than for pleasure or power, she urged participants to use education, science, and AI in ways that make us more fully human. She noted that the recording would be shared with attendees and invited all present to join the seventh webinar, on AI and Education, the following Thursday.
———
JoAnne Wadsworth is a Communications Consultant for the G20 Interfaith Forum Association and Editor of the Viewpoints Blog, IF20’s platform for commentary at the intersection of faith and global policy. She has covered the work of the Forum across multiple G20 cycles, producing summaries, articles, and editorial content that bring the voices of interfaith leaders and scholars to a broader audience. Her writing engages with topics spanning environmental governance, religious freedom, anti-racism, and the role of faith in multilateral policy processes. She is based in the United States.