OUR STORY
Four pioneering cancer centers—Dana-Farber, Fred Hutch, Memorial Sloan Kettering, and Johns Hopkins conceptualized the Cancer AI Alliance (CAIA) in October 2024. With the philanthropic and in kind support of AWS, Deloitte, Microsoft, NVIDIA, Google, and the Allen Institute for AI, CAIA is set to revolutionize cancer research through cutting-edge AI technology.
This groundbreaking collaboration harnesses the latest AI breakthroughs and unparalleled computing power to create an environment ripe for transformative discoveries. By bridging cancer centers, research disciplines, and technological advancements, we are laying the groundwork to unlock AI's full potential in the fight against cancer.
An Era of Possibility
More than 17 million people in the United States are living with cancer. While we've made significant strides in treatments—such as a 50% increase in the five-year survival rate for prostate cancer since 1975 and doubled survival rates for leukemia and myeloma—these successes must become the norm, not the exception.
Cancer in the U.S. By the Numbers
2M
People diagnosed in 2024
611,000
Cancer related deaths in 2024
69%
5-year survival rate for all cancer types
Overcoming Data Silos
The data needed to unlock AI's potential already exists but is siloed across 57 National Cancer Institutes due to patient privacy, regulatory, and intellectual property concerns. Federated learning, a recent advancement, offers a solution by allowing AI models to train based on insights derived from participating centers – without ever transferring patient data itself outside of the centers’ secure environment.
Empowering Research With Federated Learning
Federated learning enables AI models to access the necessary data on a de-identified basis without compromising privacy or security. CAIA will leverage this revolutionary approach to empower leading-edge research at participating cancer centers. Scientists and caregivers at these institutions will be empowered to think bigger and be bolder, developing, validating, and deploying new esearch breakthroughs using cutting edge technology.
Data as the Key
CAIA permits cancer centers to develop advanced AI models that could revolutionize the cancer treatment they provide to patients.Learning from the millions of people living with cancer potentially reveal genetic drivers, molecular similarities, and tumor vulnerabilities, fundamentally changing how we treat cancer.
AI's Potential
Imagine a generative AI tool specifically for cancer, allowing doctors and researchers to ask sophisticated questions and receive insightful answers. Training an AI model on de-identified cancer-specific data and decades of research could unlock new possibilities in our approach to cancer treatment.
Build to Succeed
CAIA also unites six major technology companies in this philanthropic endeavor ——AWS, AI2, Deloitte, Google, Microsoft, and NVIDIA—who have committed $65 million in resources and funding to revolutionize cancer treatment.
Fred Hutch, with its extensive experience in managing large-scale public health projects, serves as CAIA's coordinating center. Notable initiatives include:
"This alliance helps solve the key technical challenges that will enable us to securely use both AI and massive computational power to find breakthrough insights and save more lives."
— Thomas J. Lynch Jr., MD,
Fred Hutch president and director and holder of the Raisbeck Endowed Chair
The founding members of CAIA – Dana-Farber, Fred Hutch, Memorial Sloan Kettering, and Johns Hopkins – have been pivotal in cancer research, contributing to over 10 Nobel Prizes with groundbreaking discoveries like identifying cancer as a genetic disease, developing adoptive cell therapies, and even pioneering bone marrow transplants.
These institutions will enhance each other’s capabilities and attract other leading cancer centers to CAIA, expanding its invaluable stores of data.
161,000
The Women's Health Initiative, gathering health data from 161,000 women, improving prevention, and saving over $37 billion in healthcare costs.
1,300
The SWOG Cancer Research Network, a collective of 1,300 institutions, leading to FDA approval of 14 new cancer drugs.
100+
The COVID-19 Prevention Network, coordinating clinical trials at more than 100 sites, developing two effective vaccines swiftly.
Dozens
The Early Detection Research Network, bringing together dozens of institutions to accelerate early cancer detection
The founding members of CAIA – Dana-Farber, Fred Hutch, Memorial Sloan Kettering, and Johns Hopkins – have been pivotal in cancer research, contributing to over 10 Nobel Prizes with groundbreaking discoveries like identifying cancer as a genetic disease, developing adoptive cell therapies, and even pioneering bone marrow transplants.
These institutions will enhance each other’s capabilities and attract other leading cancer centers to CAIA, expanding its ability to accelerate cancer research on a large scale.
Phase 1: Building Momentum
In January 2025, leaders from CAIA's founding cancer centers and industry supporters met to begin laying the groundwork for this first-of-its-kind endeavor.
Coordinated Workstreams
We aim to have our first model ready by end of 2025. To achieve this, we are establishing working groups across the four cancer centers and our technology supporters to drive this work forward with a coordinated approach, ensuring alignment in our efforts and outcomes. Our efforts are focused on 4 key areas:
Making Data Accessible to Participating Centers
Procuring Technology
Developing and Prioritizing Use Cases
Meeting Legal and Ethical Requirements
Unified Data Standards
Though researchers themselves will never directly "see" or manipulate data from other centers in the CAIA network, the AI models they use for their research will access such data, on a de-identified basis, within the secure confines of the respective center’s environment.. For CAIA to be successful, and for researchers to work seamlessly, it is critical that the data structure and format is uniform and accessible in the same manner throughout the network. We have developed plans to harmonize the data structure and format at member centers to better unlock research insights.
