Looking back to 2006, when Google opened its first office at Taipei 101, 20 years later, Taiwan has completely transformed from its initial engineering and business base into Google's most important R&D and infrastructure hub in the Asia-Pacific region. At this significant 20th anniversary milestone, Google announced it will channel this powerful R&D momentum into the healthcare field, creating the "world's first AI Health Network."
Through in-depth cooperation with the National Health Insurance Administration of the Ministry of Health and Welfare, not only has the "diabetes AI model" been introduced on a large scale to 2 primary care clinics across Taiwan, but Google.org has also provided $100 million in funding to support rural healthcare, demonstrating that AI will be fully pushed from the laboratory to clinical practice.
20-Year Transformation: Taiwan Becomes the Core Engine of Google's Global AI Vision
From giving birth to the world's first Android phone in 2008 and the first Chromebooks in 2011, to opening the first data center in the Asia-Pacific region in the Changhua Coastal Industrial Park in 2013, Taiwan has always played a key role in Google's global innovation landscape.
With the recent relocation of its hardware R&D team to Banqiao, New Taipei City, and its AI infrastructure R&D team to Shilin, Taipei, Google Taiwan has now gathered thousands of top technical talents from over 40 countries around the world. Standing at the starting point of the next 20 years, Google Taiwan General Manager Ya-Fang Lin emphasized that the team is channeling this drive towards Taiwan's most pressing challenge—moving AI from the laboratory to the clinic to safeguard public health.
AI Enters Clinical Practice: A "Diabetes AI Model" with Efficiency Increased by 14400 Times
Google's deployment of medical AI in Taiwan has already penetrated major medical centers. For example, China Medical University Hospital was the first to adopt the MedLM model to develop precise cancer treatment plans; Chang Gung Memorial Hospital uses AI to assist in breast cancer ultrasound detection; and Taipei Medical University Hospital uses AI to automate clinical workflows to alleviate the shortage of medical staff. Recently, the National Health Insurance Administration has also developed a model based on MedGemma to assist in the large-scale analysis of more than 3 pathology reports.
The highlight of today's announcement is a major application launched under the "Community Health Plan" for primary healthcare. The National Health Insurance Administration has begun implementing the "Diabetes Artificial Intelligence Model" in 2 clinics across Taiwan. This model, based on indices certified by three major professional societies, boasts remarkably impressive efficiency improvements:
• Rapid Physician Assessment:By analyzing de-identified aggregated data for risk stratification, the time for physicians to assess a single case has been significantly reduced from 20 minutes to 25 seconds.
• Massive screening revolution:The screening of 20,000 people, which originally required 40 experts and three weeks to complete, can now be done in just 1 hour and 24 minutes with the help of AI, with the efficiency of data conversion soaring by up to 14400 times.
Health Savings Account App Imports Gemini, Millions of Dollars in Aid to Rural Areas
In addition to assisting frontline physicians, Google and the National Health Insurance Administration are also committed to bringing professional medical care into people's daily lives. It is expected that in March of this year, the "Health Savings Account" app, used by millions of Taiwanese, will officially launch a health education assistant powered by Gemini. With the user's explicit authorization and consent, this assistant can provide personalized health advice based on clinical evidence.
In efforts to bridge the digital and healthcare resource gap, Google.org, Google's philanthropic arm, announced a $100 million donation to the Digital Humanitarian Association. This funding will fully support the association.
• Integrate the diabetes AI model into its application.
• Serving patients in 300 remote community centers across Taiwan.
• Train 200 care professionals, with the goal of creating 24 health interactions.
