Artificial intelligence is reshaping healthcare, but its success depends less on scientific innovation and more on whether private healthcare systems can translate AI productivity gains into lower costs for patients. Across Asia and globally, hospitals and pharmaceutical companies are investing heavily in AI infrastructure, yet experts caution that the technology's real-world impact will be determined by business models and financial sustainability, not just technical capability.

What's Driving the Healthcare AI Investment Boom?

Healthcare organizations worldwide are deploying AI at an accelerating pace. Singapore has positioned itself as a regional hub for medical AI innovation, backed by government support through its National AI Strategy 2.0 and a S$37 billion research and innovation fund earmarked for healthcare challenges including care for an aging population. Similarly, major pharmaceutical companies are making substantial bets on AI infrastructure to accelerate drug discovery and improve clinical operations.

Bristol Myers Squibb (BMS) recently announced an expanded collaboration with Nvidia, deploying what the company describes as the most powerful and energy-efficient single-owned Nvidia computing infrastructure in the life sciences sector. This builds on a three-year partnership and reflects a broader trend across the pharmaceutical industry, with companies like Eli Lilly and Roche also expanding AI capabilities for drug discovery and diagnostics.

Local success stories demonstrate the technology's potential. KroniKare, a Singapore-based startup, developed an AI-powered wound care scanner that has been deployed across hospitals on the island, reducing wound assessment time by up to 70 percent. Yet these wins remain exceptions rather than the rule.

Why Do Most Healthcare AI Projects Fail to Reach Patients?

Despite growing investment and innovation, only a small minority of healthcare AI algorithms developed ever make it into real-world clinical settings while remaining commercially sustainable, according to Associate Professor Daniel Ting, director of the SingHealth AI Office. The biomedical sciences industry has experienced repeated funding winters and high-profile failures as investors grew cautious about the sector's profitability.

The gap between laboratory success and commercial viability reflects a fundamental challenge: scientific breakthroughs alone do not guarantee business success. As Prof. Ting explained, "If the innovation is scientifically robust and clinically applicable, the technology will remain. Then it comes down to the business model, the go-to-market strategy, and how financially sustainable it is".

Ting

"If the innovation is scientifically robust and clinically applicable, the technology will remain. Then it comes down to the business model, the go-to-market strategy, and how financially sustainable it is," said Associate Professor Daniel Ting.

Associate Professor Daniel Ting, Director of the SingHealth AI Office

How Can Healthcare Systems Translate AI Into Cost Savings?

Healthcare leaders and investors have identified several pathways for AI to deliver financial and clinical value:

• Drug Development Acceleration: Traditional pharmaceutical development takes 10 to 15 years from target identification to regulatory approval, with costs reaching several billion dollars per drug. AI could potentially cut total medical product lifecycle costs by 50 percent by helping researchers identify promising molecules, recruit suitable patients for clinical trials, and reduce costly failures.
• Hospital Operations Efficiency: Private healthcare providers including Fullerton Health and IHH Healthcare have begun deploying AI for operational tasks ranging from reading X-rays to nurse rostering and fee estimation. However, these efficiency gains must translate into lower patient costs, not just higher hospital margins.
• Hybrid Intelligence Models: BMS describes its AI strategy as "hybrid intelligence," where AI systems and researchers work together throughout scientific discovery, allowing scientists to spend less time on manual tasks and more time on complex decisions requiring human judgment.

Yet implementing these technologies carries hidden costs. As Ganen Sarvananthan, managing partner at TPG Asia, cautioned, investments in AI platforms and computing resources may initially increase overall hospital costs even as they improve productivity. The long-term return lies in enabling hospitals to treat more patients within existing infrastructure while improving outcomes, rather than simply reducing staff.

"The goal isn't speed for its own sake; it's raising the probability that each program we advance is the right one," explained Robert Plenge, Chief Research Officer at Bristol Myers Squibb.

Robert Plenge, Chief Research Officer at Bristol Myers Squibb

What Are Patients and Payers Actually Concerned About?

Rising healthcare costs have become a critical concern for patients and insurers worldwide. Dr. Prem Kumar Nair, group chief executive of IHH Healthcare, noted that medical inflation often rises faster than the consumer price index globally, leaving patients anxious about insurance affordability. Some patients report being more concerned about their insurance policies than the outcomes of their surgeries.

For AI investments to succeed, they must address this cost anxiety directly. Payers, whether governments or private insurers, are increasingly demanding that healthcare providers use technology to lower costs for consumers, not simply improve margins. This creates pressure on hospitals to demonstrate that AI-driven productivity gains translate into tangible savings or better outcomes for patients.

The pharmaceutical sector faces similar pressures. Global pharmaceutical giants including Pfizer and Eli Lilly are investing heavily in AI tools to accelerate development timelines, partly because lower development costs could eventually translate into more affordable medicines. Local biotech players like Nanyang Biologics are also betting on AI to drive long-term growth and competitiveness.

What Does Success Look Like for Healthcare AI?

The next phase of healthcare AI adoption will be defined not by technological capability but by business model viability. Healthcare systems that successfully deploy AI will be those that can demonstrate three key outcomes: improved clinical productivity, lower patient costs, and sustained profitability. This requires alignment between innovation, operations, and patient economics.

Singapore's government-backed initiatives, including the recently launched Singapore Medical Foundation AI Model (Simfoni), aim to integrate AI models tailored specifically for Singaporean patients and medical practices into public healthcare systems. These efforts signal recognition that AI's value depends on localization and integration into existing clinical workflows, not just raw computing power.

For the private healthcare sector, the challenge is steeper. Revenue-generating providers can attract better technology and equipment, but those investments must ultimately benefit patients through lower costs or better outcomes, not just higher hospital profits. As healthcare systems worldwide continue deploying AI infrastructure, the winners will be those that solve the business model puzzle, not simply the ones with the most powerful computers.