AI Reshapes Breast Cancer Screening with Personalized Risk Prediction

BOSTON, MA – April 14, 2026 – In a move signaling a major paradigm shift in preventive medicine, the National Comprehensive Cancer Network® (NCCN®) has officially updated its clinical practice guidelines to include artificial intelligence-based tools for assessing a woman's future breast cancer risk. The change, reflected in the 2026 NCCN Guidelines for Breast Cancer Screening and Diagnosis, paves the way for a more personalized and proactive approach, aiming to identify at-risk women long before a tumor develops.

This landmark update incorporates a new method of risk assessment that uses AI to analyze a woman's standard screening mammogram, identifying subtle patterns invisible to the human eye that correlate with her five-year risk of developing breast cancer. The first FDA-authorized and commercially available tool in this new category is Clairity Breast, developed by Boston-based medical technology firm Clairity, Inc.

By embracing this technology, the NCCN is addressing a critical gap in cancer screening. For decades, risk assessment has relied heavily on factors like family history and genetic mutations, such as BRCA1 and BRCA2. While crucial, these methods fail to identify the majority of women who will be diagnosed with breast cancer. This new AI-driven approach promises to cast a wider net, offering a more dynamic and individualized path to early detection and prevention for millions.

A New Standard for Risk Assessment

The updated NCCN guidelines introduce a specific, actionable metric: a five-year breast cancer risk threshold of ≥1.7% as determined by an AI-based analysis of a mammogram. Women who meet or exceed this threshold are now formally recognized as having an increased risk, triggering recommendations for more intensive surveillance.

Key changes in the 2026 guidelines include:

• Linking Risk to Action: The guidelines now directly connect this new AI-informed risk score to clinical recommendations, which may include supplemental imaging like breast MRI and the consideration of risk-reduction medications.
• Dynamic Reassessment: Recognizing that risk is not static, the NCCN calls for periodic reassessment over time, allowing a woman's care plan to evolve with her changing risk profile.
• Earlier Identification: The framework expands the identification of increased-risk individuals to begin at age 35, a critical step given the rising incidence of breast cancer in younger women who often fall outside traditional screening protocols.

This evolution represents a fundamental change in screening philosophy—moving from a reactive question of, “Do you have cancer now?” to a proactive one: “What is your risk of developing cancer in the next five years?”

“For decades, we’ve known that the mammogram contains critical information—not just about the presence of cancer, but about a woman’s future risk,” said Connie Lehman, MD, PhD, Founder and CEO of Clairity, Inc. “Advances in AI now allow us to extract that information in a clinically meaningful way. This is the foundation on which we developed Clairity Breast.”

Beyond Genes and Family History

The most profound impact of this guideline update may be for the vast population of women without known hereditary risk factors. The majority of breast cancers occur in women with no significant family history or known genetic predispositions, leaving them in a high-risk blind spot under previous screening paradigms.

“We have long relied on family history, genetic testing, and breast density to assess breast cancer risk, but these approaches fail to identify many women at higher risk,” said Robert Smith, PhD, Director of the American Cancer Society Center for Early Cancer Detection. He noted that while breast density is a known factor, it is too common and broad a measure to effectively stratify individual risk. “This situation highlights a fundamental gap in our current approach, where AI-based analysis of mammograms represents an important new direction to overcome these limitations.”

Clinical studies have shown that AI algorithms can outperform traditional clinical risk models. Research has demonstrated that AI models trained on hundreds of thousands of mammograms are more accurate at predicting five-year breast cancer risk than established tools like the Breast Cancer Surveillance Consortium (BCSC) model. The power of these AI tools lies in their ability to analyze the entire mammogram at a pixel level, identifying complex textural and structural patterns that signal future malignancy.

“It’s encouraging to see advances in breast cancer risk assessment beginning to reach clinical care, including AI-based approaches that may help identify higher-risk women earlier—particularly those under 50 who might otherwise go unflagged,” said Dr. Judy Garber, Breast Cancer Research Foundation Scientific Director.

The Technology and Path to Practice

Clairity Breast secured its place as a pioneer in this field after receiving a De Novo authorization from the U.S. Food and Drug Administration (FDA) in June 2025. This first-in-class designation created a new regulatory category for imaging-based AI models designed for future risk prediction. The tool was validated using approximately 77,000 mammograms from diverse geographic and demographic populations to ensure its generalizability.

While Clairity is the first to have its technology included in NCCN guidelines, the competitive landscape is heating up, with companies like Therapixel also receiving FDA authorization for similar risk prediction tools. This signals a broader industry trend toward leveraging AI not just for detecting existing cancer but for predicting its onset.

The inclusion in national guidelines is a crucial step toward widespread clinical adoption. “This is a meaningful evolution in how we think about breast cancer risk,” stated Beth Mittendorf, MD, PhD, Chief of Breast Surgery at Beth Israel Deaconess Medical Center. “Incorporating this approach into national guidelines expands our ability to identify women who may otherwise not be recognized as being at increased risk.”

However, the path from guideline to universal practice faces hurdles. Experts note that clinical adoption of AI tools can be slow, hampered by a need for further real-world validation, provider education, and seamless integration into existing radiology workflows. Questions surrounding cost-effectiveness and reimbursement policies from payers will also need to be addressed, though modeling studies have suggested that AI-stratified screening could ultimately be cost-saving for health systems.

For this technology to truly transform patient outcomes, it must be implemented equitably. “This provides an opportunity to translate these advances into clinical practice in a way that ensures women truly benefit,” said Tari King, MD, Chief of Breast Surgery at Emory University School of Medicine. “That means making risk assessment accessible, understandable, and actionable across care settings so that more women can receive care tailored to their individual risk.”

The new framework is expected to heavily influence clinical practice, payer policies, and health system strategies, accelerating the momentum toward a future of precision prevention where data-driven insights are used to tailor screening, reduce the burden of late-stage disease, and ultimately save more lives.