AI could decide what you pay for insurance, based on data that you DIDN’T know you shared…

24 Aug 2026 · 1 min

Watch on YouTube

0:56

In this monologue, Dan Novaes explains how traditional insurance pricing relies on broad averages such as age, location, and driving history, but AI can shift the model to individualized risk assessment by processing massive amounts of personal data. He highlights the potential value for insurers in achieving more precise pricing, while raising ethical and regulatory concerns about the types of data that could be used—shopping habits, fitness trackers, online behavior—and the opacity of algorithmic decisions. Novaes emphasizes that the core debate will center not on AI's predictive power, but on which aspects of a person's life should be permissible for insurers to analyze, prompting increased regulatory scrutiny of automated decision‑making.

Chapters

Insurance has always worked on averages— your age, your location, your driving history. Then companies estimate how risky you are. AI could make that much more precise. Instead of putting millions of people into broad categories, AI can analyze enormous amounts of information and predict risk at an individual level. For insurers, that's incredibly valuable. But here's where things get controversial. What information should they be allowed to use? Your shopping habits, your fitness data, your online behavior— even information that seems unrelated could potentially reveal patterns about risk. And that creates a problem. An algorithm algorithm might decide you're expensive to insure without you fully understanding why. That's why regulators are paying closer attention to automated decision-making— because AI could make insurance more accurate and potentially more personalized. But it could also create a world where companies know far more about your behavior than you're comfortable sharing. The real debate won't just be about whether AI can predict risk; it'll be about which parts of your life it should be allowed to analyze.