Predicting Cancer Before It Starts: An AI Milestone in Women’s Health

I don’t usually post about client work here, but this breakthrough is too important not to share. I'm proud of what we built together, and even more proud of what it means: this will save lives.

For over 60 years, mammograms have helped catch cancer early. But they’ve never been able to tell someone who appears perfectly healthy that they may be at higher risk in the future. That changes now.

Clairity partnered with KUNGFU.AI to build and train a groundbreaking AI model on over half a million mammograms. The result is the first FDA-authorized AI platform that can predict a woman’s five-year risk of developing breast cancer using only a standard screening mammogram.

To be clear, this model doesn’t detect cancer. It predicts risk before any signs of cancer appear. This is a leap beyond traditional detection tools, and far more challenging to pull off.

We validated the model on 77,000 mammograms from five different sites across the country, each with real five-year patient outcomes. It delivered a step-function improvement in predictive accuracy, something the field has been chasing for decades.

And it isn't just more accurate. It's also significantly less biased, making consistent risk predictions across diverse populations. In a field where equity in access and accuracy is a known problem, that’s a huge deal.

Think of it like this: instead of using a smoke detector to catch fires when they’ve already started, this model gives us a chance to identify which homes are more likely to catch fire in the first place—so we can act sooner.

This is the kind of impact AI should be making in healthcare. Real science, real validation, real-world benefit.

I'm incredibly proud of the KUNGFU.AI team and grateful to work with partners like Clairity, who are setting a new bar for what’s possible. Check out the full case study here.