The University of Hawaii's Cancer Center and John A. Burns School of Medicine are building a new artificial intelligence data center with more than $12 million in National Institutes of Health funding, aiming to sharpen cancer detection and treatment for diverse populations across the Pacific islands. The five-year initiative will outfit a garage-sized room at the Cancer Center in Kakaako to power AI research that accounts for the region's unique ethnic makeup-a factor often missing in current diagnostic models.
Building a Pacific-first AI research hub
The funding will convert a ground-floor, air-conditioned space at the Cancer Center into Hawaii's first AI-driven data center of its kind. Renovation is already under way, with the facility expected to be operational by the end of the year. Once complete, it will accelerate work flowing from the new Pacific Center for Artificial Intelligence and Data Science in Medicine, co-led by John Shepherd, the UH Cancer Center's chief scientific officer, and Youping Deng, a JABSOM cancer researcher.
The project draws roughly $300,000 for the physical conversion and high-speed computing equipment. U.S. Senator Brian Schatz, who pushed for the federal grant, called it "very good news for Hawaii and the Cancer Center and School of Medicine" that will assist physicians, healthcare workers and researchers throughout the region. "There is not a comparable research institution in the Pacific," Schatz said. "The quality of the research and the sophistication of the research and the size of the grant means that Hawaii is going to lead in this space for years."
Closing the ethnicity data gap in medical AI
Existing AI databases already allow clinicians to photograph suspicious moles and receive automated analysis-sometimes avoiding an unnecessary biopsy. But current models rarely account for the differences among the ethnic backgrounds found in Hawaii and across the Pacific, Shepherd said. Grouping Filipino, Korean, Japanese and Chinese patients under a single "Asian" label can mask clinically important variation in conditions like skin cancer.
Researchers plan to gather and share detailed, ethnicity-specific data that mainland and Pacific-region providers can access. The more granular information could help clinicians distinguish between groups that today tend to be lumped together. Healthcare workers interested in integrating such tools into practice can pursue specialized education through programs like AI for Healthcare Courses.
Workforce and regional impact
Shepherd, 63, acknowledged AI is "a young man's game. But I really want to stay in the game. I want to stay relevant." The data center will need roughly 26 people, primarily with computer science and programming backgrounds, and could draw ex-pats back to Hawaii for high-skill roles.
Schatz estimated the initial investment could spin off an additional $50 million to $100 million in research funding over time. He also pushed back on fears of a mainland-style, warehouse-sized facility consuming outsized energy and water. "There's zero chance there will be a data center in Hawaii that size and they're just too expensive," he said. Shepherd left the door open for a larger center someday, pointing to local renewable sources: "We've got plenty of sun, we've got plenty of chill water in the ocean. We've got the volcano for power. We've got wind. Ultimately, we don't necessarily need to use more of our imported oil and fossil fuels."
He compared AI's arrival to a historical warning from Native Hawaiian historian John Papa 'Ī'ī about Westerners coming "like a tsunami." Shepherd said, "AI is a lot like that. It's a literal tsunami and we can either choose to ignore it and have it impact us in a random way, or we can choose how we're going to meet it so that it serves us."
Why this matters for healthcare professionals
The new data center will build datasets that reflect the Pacific's ethnic diversity-information clinicians can eventually use to make better diagnostic decisions for patients who are often underrepresented in AI training data. As the center's research becomes available, professionals who learn to work with AI-assisted diagnostic tools will be positioned to apply those insights directly in clinical settings, particularly when treating Asian, Native Hawaiian, and Pacific Islander populations whose health patterns current models often miss.
Your membership also unlocks:
