Working at Color, it seems like every day I have a conversation that dovetails into a story about an unexpected cancer diagnosis.

“I didn’t know I should get screened early – my doctor never told me that.”

“I kept having to reschedule my mammogram because of work.”

“She couldn’t get in for six months, and what would’ve happened if that appointment could’ve been sooner?”

These are real experiences that underscore a critical gap in how our healthcare system approaches prevention and early detection.

The gap in breast cancer screening

Why does this keep happening?

In breast cancer, screenings are covered at 100% for most women who meet guidelines. Each October, we see just how much advocacy, education, and conversation exists around breast cancer, screening, and survival rates.

And yet, up to 1 in 3 women remain behind on their mammograms. Millions more women in the U.S. are at higher risk over their lifetimes and do not know that they should be screened earlier or more frequently. 

Rethinking access with Google Cloud

This is why at Color we are working to rethink the cancer care model for screening and prevention, and why we have worked with Google Cloud and Google.org on an initiative to support getting tens of thousands of women screened, kicking off in Breast Cancer Awareness Month and continuing through the end of the year.

You can get started on screening here. 

Together, we’re using technology to make screening easier to access, starting with one of the biggest barriers: clinical capacity. 

Where AI makes the difference

This is possible in part because of an AI agent we built to help expand access to mammograms by tackling one of the most common chokepoints in screening availability: clinical capacity to identify eligible individuals and order screens.

This isn’t a problem many think about—usually the orientation in both AI and overall health policy is to consider how to expand the supply of clinicians, reduce administrative burden, or focus on coverage. 

The promise we see in AI is to expand access to expertise. Getting a mammogram done may seem quite simple among all the tests/screens out there, but it first requires:

• A patient to see and understand the need for screening (Why this is important to me, now)
• A history collection to understand eligibility (Can they start screening younger due to risk? Do they need a breast MRI as well, due to dense breasts? When was their last screening?)
• Order generation (A clinician ordering a mammogram)

An AI-based interface for these steps helps patients start with us in a way that is independent from clinician time and capacity. The method of information discovery—both to the patient and to the clinician—doesn’t rely on a phone call or a visit that takes up valuable, expensive time.

Making time for what matters most

Think about the last time you went through your personal or family health history in a visit with your primary care doctor: In the best of circumstances, it was likely only a few minutes long. And in all likelihood, there was not time for any back-and-forth to work through concerns about screening:

“Why is it important?”

“Will it hurt?”

“I’m not sure I want to learn if I have cancer, though?”

Today’s healthcare model leaves no space to understand important details about each patient or to fully address patients’ questions.

This approach allows Color’s clinical team to gather and evaluate rich, comprehensive information about each patient’s situation: unique symptoms and timelines, clinical considerations like recent changes to the breast, personal risk factors, and detailed family history, in the user’s own words. The agent asks what’s relevant, then turns those nuanced details into a clear, structured assessment for our clinicians to review—in seconds.

Looking ahead

This application of AI—accessible, abundant clinical expertise—is just as big an opportunity in AI in healthcare as any of the earliest applications we are already seeing. It speaks to both patients and clinicians. And it gets a more comprehensive view of clinical need than traditional models do.

The future here is one that could have a massive impact on the increasingly untenable cost of healthcare in this country.

We hope this initiative with Google Cloud and Google.org can help you or a woman in your life get screened this year.