Color Health receives Newsweek AI Impact award for “AI Healthcare: Best Outcomes, Diagnostics”

Award goes to Color’s Cancer Copilot, an AI-based diagnostic tool, implemented with real-world patient cases at a major health center

When we set out to build an AI implementation for healthcare, specifically Cancer Care, we just hoped it would work. Our thesis was that if you increased the abundance of a healthcare component more scarce than any other—expertise—it would be a huge win for patients. So we built Cancer Copilot, a tool that expands access to oncology expertise to any clinician, no matter the location.

In recognition of this work, we’re proud to receive Newseek’s AI Impact award for “AI Healthcare: Best Outcomes, Diagnostics,” announced today. This validates the team’s efforts in building AI that has the potential to change care delivery, going beyond the scribe approach of notetaking and visit summaries, to provide cancer patients the best possible care, and shrink the time it takes to go from diagnosis to treatment.

As a clinical tool, we were extremely committed to ensuring our Cancer Copilot was creating something highly accurate and avoiding hallucinations. To achieve this, we leveraged a leading large language model (LLM), and wrapped the technology in machine learning steeped in condition- and institution-specific cancer treatment guidelines. This ensured the LLM would be on its best behavior, and only recommend next steps based on appropriate, real-world guidelines – not hallucinations. 

Working with the University of California, San Francisco (UCSF) to implement the Cancer Copilot, Color achieved >95% accuracy on diagnostic recommendations for cancer patients. We also identified critical gaps in care for patients to address before beginning treatment, and reduced the time it takes clinicians to conduct these workups to roughly ten minutes, down from what traditionally takes ~2 hours. Another key element of our approach is allowing for clinician intervention throughout the recommendation process, and exposing all of the Copilot’s logic used to make decisions, bringing the AI out of the black box and, well, into the clinic.

By expanding oncology expertise beyond major cancer treatment centers, and reducing the time it takes clinicians to assess next steps for cancer patients, we can get patients into treatment much more quickly, which is crucial given every month of treatment delay increases the risk of mortality by up to 13%.

In addition to the Newsweek award, Color’s healthcare AI efforts were featured in an abstract at the 2025 American Society of Clinical Oncology (ASCO) Annual Meeting this month. ASCO is the seminal oncology research conference, and Color’s was one of a few AI applications recognized as directly improving cancer care delivery.

We are continuing to expand our applications beyond diagnosis and into stronger care models for personalized cancer screening, risk prediction, and more. If you’re interested in learning more or joining us, please contact us at copilot@color.com.