As the use of artificial intelligence (AI) expands, scientists are exploring ways to integrate this technology into breast cancer research and improve breast cancer care.  

In this episode of Breast Cancer Breakthroughs, we speak with Komen Scholars, Nancy Lin, M.D. and Christina Curtis, Ph.D., along with Komen Metastatic Breast Cancer (MBC) Steering Committee member, Megan Hanvey, to discuss how AI is being used in breast cancer research and care today, its potential for the future and what patients should know as they navigate today’s AI-driven landscape. 

AI in Breast Cancer Care & What is Possible in the Future 

In 1998, AI was introduced into breast cancer care with FDA-approval of the first computer-aided detection for mammography. AI tools have steadily improved since their inception, and today, radiologists may use AI to help them read mammographic images more quickly and accurately. 

Dr. Lin says that the use of AI can be placed in one of two categories: “to do work that can be done by a human”— such as reading mammographic images — or “to do work that can’t be done by a human…and, I think that is really the wave of the future.” 

Dr. Lin outlined questions that currently elude radiologists but AI may be able to reliably answer in the future. “Can AI not only say whether there is cancer or not in this image, but what is this specific person’s risk of developing breast cancer over the next five years?” she asks. “And if it’s high enough, should that person have additional imaging beyond mammograms? Or, based on an image, does this person have a higher or lower risk of cancer recurrence? Finally, if we’re looking at a radiology scan, is the patient benefiting or not benefiting from treatment?” 

We are starting to see some of these advancements now. In 2025, the FDA granted authorization to the first-ever AI-powered tool to predict a the risk of developing breast cancer over the next five years. More recently, in 2026, the FDA cleared the Multimodal Artificial Intelligence (MMAI) platform to predict long-term outcomes and potential therapy benefits using histopathology images (microscopic photographs of tumor samples) and clinical variables for individuals with early-stage, HR+, HER2-negative invasive breast cancer. 

Looking beyond just scans and images, could AI sift through an array of medical data to uncover critical insights into how breast cancer begins and how it behaves? Dr. Curtis considers this an exciting future possibility. “We’re starting to have the capability to link the tumor images to the genomics to the electronic health records,” she said. “So, this notion of multiple modalities [pulling together and learning from different types of data] is very powerful. We can combine them and potentially improve our understanding and our predictive ability.” 

Considering the possibility of using AI to identify the optimal personalized treatment plan, Dr. Lin adds, “Could we one day have AI which combs through a million people’s worth of de-identified [electronic health] records to determine the sequence of therapy most likely to give [an individual] benefit? I don’t think we are close to that yet, but that would be a game changer.” 

What is Limiting AI Expansion? 

So, what is limiting AI tools from accomplishing all these possibilities? Dr. Lin and Dr. Curtis both agreed on one answer — data. 

“One of the most important things is you must have good quality data coming in so that the AI is trained on reality,” says Dr. Lin. “And for certain things — like a mammogram and whether cancer was diagnosed or not — the data are very, very good.” 

Data from treatment plans following a diagnosis are much more complex. “An important question in practice is: What is the best sequence to give treatments in? That would be an incredible thing to train AI to try to learn. But to do that, you need high-quality data to train a model,” says Dr. Lin. “Those kinds of data exist, but I don’t think they’re of the quality that will allow us to answer those questions the way we want to answer them.” 

Dr. Curtis adds, “The predictions we make — the things that we learn using AI — are only as good as the data we input. So, we need diverse data sets to input [when training an AI tool]. That means data that is representative of diverse patient populations. It means different types of data across different stages. And I would say that data is what can limit advancing the most powerful class of AI models.” 

Patient Perspective on AI — To Use or Not to Use 

AI is not just a tool used by clinicians and scientists. Patients and their families can leverage these tools to help educate themselves on their diagnosis, treatment plan, side effects and overall health. 

Megan was first introduced to the idea of AI by a friend. Like many, she debated using it because of concerns about data privacy. “I finally surmised that I was okay with it,” she says. “I have stage 4 [breast] cancer, and it felt worth it to me to take that risk.” 

The AI tool Megan used was designed to assist cancer patients. It helps her organize and simplify the mountain of health information available so she can zero in on key topics to discuss with her oncologist. “You turn to the internet, and there’s so much information. And it can be really overwhelming and scary,” Megan says. “The tool that I use really helps to focus on my specific tumor mutations and my specific profile. It is also important to me to use an AI tool built from trusted sources, like medical journals.” 

The AI tool primarily helps Megan prepare for appointments, which allows her to maximize time with her oncologist. “It helps me generate questions [for my oncologist] and to think a little bit further about things that I wouldn’t have thought of on my own without a medical background. I also use the tool with my oncology team to find ways to manage side effects from treatment.”  

Megan notes that living with metastatic breast cancer requires balancing survival with quality of life. Along with her medical team, the AI tool helps her achieve that balance. “I feel more confident and empowered,” she shares. 

Regardless of whether you choose to use AI tools, we encourage caution when sharing personal details on these platforms. Always use trusted sources, like komen.org, and consult your doctor when making medical decisions rather than relying on AI. You can learn more about how to navigate these tools safely and effectively in our blog, The AI Toolbox: How to Use AI as a Resource. 

Sign up to stay informed about future episodes of Breast Cancer Breakthroughs; we will delve into the most exciting advances in breast cancer treatment and technology coming soon for patients.  

Content covered in the Breast Cancer Breakthroughs educational series may be an emerging area in research or technology. This information is being provided for educational purposes only and is not to be construed as medical advice. Talk with your doctor about what is right for you. 

For more information on clinical trials or if you need support as you go through treatment, the Komen Patient Care Center can help. Please contact the Komen Breast Care Helpline at 1-877 GO KOMEN (1-877-465- 6636) or email helpline@komen.org.               

Find a clinical trial. 

Let us know what you think of this content as part of the Breast Cancer Breakthroughs Series.   

Breast Cancer Breakthroughs is possible thanks to our series sponsors: