When Your Sleep Data Costs You a Mortgage

Simon Carver

Monday morning. David opened an email expecting, well, expecting a congratulations letter, but instead his mortgage refinance was just... flat out denied. His credit score was a near-perfect 812. His salary had just gone up, he'd never, not once, missed a payment on anything. But the- the loan officer just kind of shrugged over email and said, um, the model determined your long-term risk profile no longer met our lending criteria. Nobody mentioned his smartwatch. Nobody mentioned that sleep study he did last autumn, or the diabetes screening he took at a workplace wellness day. Nobody had to.

Chris J. Murphy

Because the data was already in the room with them. I- I mean, Simon, this is what people don't get. We think of our medical history as this locked vault, you know, protected by federal laws like HIPAA here in the US. But the vault has a massive back door. We're talking about consumer data brokers taking completely legal, unregulated crumbs of your life—what you buy, how many hours your watch says you slept, whether you bought a continuous glucose monitor on a whim—and weaving them into a shadow score.

Simon Carver

It- it is a shadow score. That is the perfect way to put it, CJ. And welcome to the show, by the way! For everyone listening, I'm Simon Carver, and joining me today is the brilliant Chris J. Murphy—CJ—co-author of The Last Job You'll Ever Hate and a true voice of sanity in this wild tech transition. If you're finding this episode eye-opening, do us a quick favor: hit that subscribe button, share this with a colleague or a friend who has a smartwatch on right now, and leave us a review. It really helps us keep these human stories going. But CJ, back to this back door. How on earth is a bank allowed to use sleep patterns to deny a mortgage? Is that even legal?

Chris J. Murphy

It's entirely legal because they aren't looking at "medical records." They're buying predictive risk models built from consumer behavior. If you buy unscented lotion, certain vitamins, and a specific brand of supportive shoes, a machine learning model doesn't need a doctor's diagnosis to flag you as a potential diabetic or a high cardiovascular risk. It's not clinical certainty, Simon. It's correlation. And in the financial world, correlation is cheap, fast, and incredibly lucrative to act on.

Simon Carver

So we've shifted from, um, looking at what actually happened—your actual history of paying bills on time—to this weird, algorithmic fortunetelling. We're pricing people's futures based on a probability of whether they might get sick five years from now. I mean, think about the workforce. You apply for an executive gig, you're the top candidate, but some HR predictive tool flags you as a high risk for future absenteeism because your wearable data shows your sleep quality has been slipping. They don't tell you that, of course. They just choose another candidate.

Chris J. Murphy

Exactly. The real danger isn't that the AI is malicious; it's that leaders are trusting these statistical forecasts more than they trust the actual human being standing in front of them. It's a massive governance failure. If you're a leader running these models, you have to actively audit them for what we call proxy bias. If your algorithm starts penalizing people who live in food deserts because their grocery receipts don't show enough organic kale, you've just automated discrimination under the guise of "risk management." We have to demand explainability. If a model makes a decision about a human life, a human must be able to explain exactly why.

Simon Carver

Yes! And as individuals, we've got to treat our personal data like a financial asset. Don't just blindly click "accept" on those company wellness programs without reading the privacy disclosures. Be incredibly mindful of what data your wearables are broadcasting into the cloud. We have to draw a line. Because at the end of the day, an algorithm can estimate probabilities, but it can't measure human resilience. It can't see the determination of someone who refuses to be defined by a risk score. That's what the Human Workforce is all about—keeping humanity in control of the decisions that actually matter. If this resonated with you, please subscribe, share the episode, and let's keep this conversation going. Thanks for being here, CJ.

Chris J. Murphy

Always a pleasure, Simon. Take care of your data, everyone.