New Flourishing AI Benchmark Measures Performance of Top LLMs Across Key Dimensions of Human Well-Being

Research-backed framework measures how well top AI LLMs support human flourishing and invites community collaboration to help build standards that support holistic well-being

BOULDER, Colo., July 10, 2025 /PRNewswire/ -- Gloo, a technology platform serving the faith ecosystem, today released the Flourishing AI (FAI) Benchmark, a new research-based framework designed to evaluate how well today's leading Large Language Models (LLMs) support human flourishing. Created in collaboration with researchers from Gloo, Barna Group, and Valkyrie Intelligence, the research draws upon the published research from the Global Flourishing Study and is designed to serve leaders and organizations committed to human flourishing. The FAI Benchmark represents one of the first comprehensive assessments of AI values, measuring not just technical capabilities, but how well models support Character, Relationships, Happiness, Meaning, Health, Finances and Faith.

"Core to what we do at Gloo is to serve those organizations who are helping people flourish," said Scott Beck, CEO at Gloo. "There is an urgent need to shape technology for good — and these standards and benchmark measures are an important way we can serve the faith and flourishing ecosystem."

The FAI Benchmark is the start of a collaborative, open process for exploring the impact of AI on holistic human well-being. Gloo is publicly releasing the methodology and invites researchers, developers and interdisciplinary experts to contribute to an evolving standard for values-aligned AI.

The FAI Benchmark consists of over 1,200 curated questions drawn from academic research, professional licensing exams, existing LLM benchmarks and more. The evaluation uses both objective and subjective questions, employing multiple judge LLMs with specialized expert personas for each dimension. Model responses are judged based on both their primary dimension's rubric as well as other dimensions that the judge LLMs deem relevant.

"Throughout my career, I've helped set industry benchmarks to make technologies more robust, transparent and effective," said Pat Gelsinger, executive chair and head of technology at Gloo. "AI is one of the most consequential technologies shaping humanity's future. To guide its development, we must measure it against the ultimate standard — human flourishing. Today, that vital journey begins."

Key findings from the initial FAI Benchmark:

• No current model achieved a 90-point threshold for excellence in supporting human flourishing across all seven dimensions. Average model scores across individual dimensions include Character (58), Relationships (67), Happiness (65), Meaning (56), Health (72), Finances (81) and Faith (35).
• Holistically top-performing models include OpenAI o3 (72 points), Gemini 2.5 Flash (68), Grok 3 (67) and GPT-4.5 (66).
• The greatest performance gaps were found in Faith and Meaning, suggesting the need to further develop model capabilities in areas such as existential reasoning, ethical reflection and virtue-based considerations.
• Models performed best in areas like Finance and Health, with several systems such as OpenAI o3, Grok 3 and OpenAI GPT-4.1 achieving their highest scores in these categories.

The findings reveal a critical gap in AI "wisdom." While models can handle practical questions about finances and health, they consistently struggle to provide meaningful guidance on life's deeper questions, particularly related to faith, purpose and moral character. This suggests that current AI development prioritizes technical capabilities over the kind of wisdom humans need for truly flourishing lives.

"The FAI Benchmark offers a new lens for evaluating AI — not just in terms of usefulness or safety, but in how well it supports the full spectrum of human flourishing," said Steele Billings, chief AI officer at Gloo. "By making our framework open and collaborative, we hope to inspire the broader AI research community to join us in shaping AI that genuinely empowers people to grow and thrive."

Through this work, Gloo invites interdisciplinary researchers — from psychology, philosophy, theology, ethics, and AI safety — to engage with and refine this foundational framework. Offering open data, transparent evaluation protocols and a community-driven model, Gloo is seeking peer review and collaborative inquiry into future research directions, including question iteration and improvement, rubric evaluation, open source repositories and more. Full methodology details are available in the downloadable research paper below.

The FAI Benchmark will be updated regularly as new AI models are released and as new research about human flourishing is completed. For more details, visit https://gloo.com/fai.

About Gloo
Gloo is a leading technology platform for the faith ecosystem, providing values-based AI, resources, insights and funding so people and communities flourish and organizations thrive. Gloo serves over 100,000 faith, ministry and nonprofit leaders and is based in Boulder, Colorado. For more information, visit Gloo.com

SOURCE Gloo