How Contactless Vitals Reduce Underwriting Fraud

Insurance fraud costs the U.S. industry an estimated $308.6 billion per year, according to the Coalition Against Insurance Fraud. Life insurance underwriting sits right in the middle of that problem. When an applicant lies about their smoking status, hides a hypertension diagnosis, or omits a heart condition, the insurer doesn't find out until a claim gets filed — sometimes years later, sometimes never. Contactless vitals technology, which uses a smartphone camera to measure heart rate, respiratory rate, and blood pressure indicators in real time, creates a new layer of objective data that makes certain types of fraud much harder to pull off.

"All three digital underwriting evidence sources demonstrated significant value, both individually and in combination." — RGA, "Assessing Mortality Impact of Digital Underwriting Evidence" (2025)

Where underwriting fraud actually happens

Most people think of insurance fraud as staged car accidents or fake disability claims. Underwriting fraud is quieter. It happens at the application stage, when someone misrepresents their health, lifestyle, or medical history to get a lower premium or avoid being declined entirely.

The most common forms in life insurance:

• Omitting diagnosed conditions like diabetes, hypertension, or heart disease
• Lying about tobacco or nicotine use
• Understating alcohol consumption
• Failing to disclose prescription medications
• Using someone else's identity for the application

The two-year contestability period gives insurers a window to investigate and deny claims if misrepresentation is discovered. After that window closes, the policy generally becomes incontestable unless outright fraud can be proven. The problem is that many misrepresentations only surface when a claim is filed, and by then the insurer has been collecting premiums based on incorrect risk assumptions for years.

Traditional underwriting caught some of this through paramedical exams — a nurse would take blood, urine, and blood pressure readings, and those results would either confirm or contradict the application. But as the industry moves toward accelerated and fluidless underwriting, that physical checkpoint disappears. Electronic health records, prescription histories, and claims data fill part of the gap. They don't fill all of it.

The gap between self-reported data and objective measurement

Here's the core problem accelerated underwriting creates for fraud detection: the data sources it relies on are backward-looking. EHR records tell you what a doctor documented at the last visit. Prescription data shows what medications someone filled. Claims data shows what procedures were billed. None of these tell you anything about the applicant's current physiological state at the moment they're applying.

Someone who was diagnosed with hypertension three years ago but stopped seeing their doctor has clean recent records. Their blood pressure could be dangerously elevated right now, and no amount of EHR scraping would reveal it. A smoker who buys cigarettes with cash and hasn't had a recent checkup won't show nicotine-related diagnoses in their health records either.

Fraud type	Traditional FUW detection	Accelerated UW detection	Contactless vitals detection
Undisclosed hypertension	Blood pressure reading at paramedical exam	May appear in EHR if recent visit	Elevated BP indicators captured in real time
Hidden tobacco use	Cotinine in blood/urine sample	Rx for cessation aids (if prescribed)	Elevated resting heart rate, HRV patterns consistent with nicotine use
Omitted cardiac condition	ECG or lab abnormalities	Diagnosis codes in EHR/claims	Irregular heart rate patterns detected during scan
Identity fraud	Nurse verifies applicant in person	No physical verification	Liveness detection via rPPG pulse signal
Concealed stress/anxiety disorders	May appear in lab work	Rx for SSRIs, benzodiazepines	Stress indicators and HRV abnormalities

The contactless vitals column isn't theoretical. Camera-based photoplethysmography captures blood volume changes through facial skin, and those signals carry information about cardiovascular and autonomic function that self-reported data simply can't provide.

How rPPG-based vital signs capture works

Remote photoplethysmography, or rPPG, works by analyzing subtle color changes in facial skin caused by blood flowing through capillaries. Every heartbeat pushes oxygenated blood to the face, and a smartphone camera can detect those micro-fluctuations in reflected light. From that signal, algorithms extract heart rate, heart rate variability, respiratory rate, and blood pressure estimates.

The applicant opens the insurance application on their phone, the camera activates for 30 to 60 seconds, and the scan captures a snapshot of their cardiovascular state. No equipment, no appointment, no nurse. The data is generated at the moment of application.

For fraud detection, this matters for two reasons.

First, you get objective physiological data that the applicant can't easily fabricate. You can lie on a questionnaire. You can avoid going to the doctor so there's nothing in your EHR. You can't fake a normal heart rate variability pattern when your autonomic nervous system is compromised by untreated hypertension or chronic nicotine use.

Second, rPPG provides liveness detection as a byproduct. The same blood flow signal that measures heart rate also confirms that a living person is sitting in front of the camera. Research published in the ACM International Conference on Multimedia (Li et al., 2019) demonstrated that rPPG-based liveness detection can distinguish real faces from photos, videos, and 3D masks by detecting the presence or absence of genuine blood flow signals. A synthetic face or a deepfake video won't produce the micro-color fluctuations that a real pulse creates. This addresses identity fraud at the application stage without requiring the applicant to perform awkward head turns or blink sequences.

What the research says about biometric fraud indicators

The connection between resting physiological metrics and health behaviors is well documented in clinical literature, and it gives contactless vitals fraud-detection potential that goes beyond what a simple "are they alive?" check provides.

Heart rate variability and smoking

A 2018 meta-analysis published in the journal Nicotine & Tobacco Research by Tobaldini et al. at the University of Milan found that chronic smokers show significantly reduced heart rate variability compared to non-smokers. HRV is a measure of the variation in time between consecutive heartbeats, and lower HRV indicates reduced autonomic nervous system flexibility. When an applicant claims to be a non-smoker but their HRV pattern looks like someone with chronic nicotine exposure, that's a data point worth investigating.

This won't replace a cotinine test in terms of certainty. But in an accelerated underwriting flow where there is no cotinine test, it provides a signal that would otherwise be missing entirely.

Resting heart rate and cardiovascular risk

Elevated resting heart rate has long been associated with increased cardiovascular mortality. A study by Jensen et al. published in Heart (2012) tracked over 2,798 men for 16 years and found that resting heart rate above 80 bpm was associated with substantially higher all-cause mortality risk compared to rates below 50 bpm. An applicant who claims to be in good cardiovascular health but scans with a resting heart rate of 95 bpm warrants additional scrutiny — or at minimum, a flag for the underwriter to pull additional evidence.

Blood pressure estimation

Camera-based blood pressure estimation is still an evolving field, and accuracy varies across implementations. But the trend lines are clear: multiple research teams, including work by Luo et al. at the University of Toronto (2019), have demonstrated that transdermal optical imaging can estimate blood pressure from facial video with reasonable correlation to cuff-based readings. Even imprecise estimates have underwriting value. If the estimate suggests stage 2 hypertension and the application says "no history of high blood pressure," that discrepancy triggers a closer look.

Fraud detection at scale without adding friction

One reason underwriting fraud persists is that the traditional tools for catching it are expensive and slow. Ordering an attending physician statement costs $50 to $200 and takes weeks. Paramedical exams add cost and delay that carriers are actively trying to eliminate because they drive applicant dropout. Gen Re's 2024 survey found that the average completion rate for accelerated underwriting programs exceeds 80%, compared to traditional processes where paramedical exam scheduling alone causes significant falloff.

Contactless vitals scanning adds a fraud detection layer that takes under a minute and requires nothing from the applicant except looking at their phone. The marginal cost is effectively zero compared to lab work or nurse visits. And because it's embedded in the digital application flow, it doesn't create the friction that causes applicants to abandon the process.

This matters for carriers trying to balance two competing priorities: reducing fraud exposure and maintaining conversion rates. Every additional step in the application process loses applicants. A sub-60-second camera scan that runs inside the existing application doesn't create a new step — it augments an existing one.

Industry applications

Accelerated underwriting programs

Carriers running accelerated underwriting programs without paramedical exams lose the single best source of real-time health verification they had. Contactless vitals partially restore that capability without reintroducing the delays and costs that made carriers move away from paramedics in the first place. Munich Re's 2024 U.S. Individual Life Accelerated Underwriting Survey found that maximum face amounts have climbed to $2.5 million for accelerated programs. At those amounts, the cost of undetected fraud at application stage is significant.

Group enrollment screening

Group life and voluntary benefits programs process large numbers of applicants quickly, often with simplified underwriting that relies almost entirely on self-reported health information. Adding a camera-based vital signs scan during digital enrollment provides a layer of biometric verification that is otherwise absent in the group context. A study by the Society of Actuaries (2023) on group underwriting trends noted that simplified issue programs accept higher baseline mortality risk in exchange for volume and speed. Contactless vitals could help group carriers identify the outliers in that risk pool without slowing enrollment.

Direct-to-consumer and embedded insurance

The embedded insurance model — where coverage is offered at point of sale through a retail or fintech partner — operates on pure digital rails with minimal underwriting. Fraud risk in these channels is higher because there's less verification by design. A camera scan during the purchase flow provides a biometric anchor that the current embedded insurance process lacks entirely.

Current research and evidence

The body of research supporting camera-based vital signs measurement has grown substantially. A systematic review by Hassan et al. published in Sensors (2022) cataloged over 100 studies on rPPG-based vital sign measurement across various populations and conditions. The technology has moved past proof-of-concept into practical deployment across telehealth, clinical trials, and now insurance.

For anti-spoofing specifically, Kim et al. (2022) published work in Sensors demonstrating a face biometric spoof detection method using rPPG signals that achieved high accuracy in distinguishing real faces from presentation attacks. The authors noted that while replay attacks remain a challenge, combining rPPG liveness with other signals (like device attestation and behavioral biometrics) creates a multi-layered defense.

The insurance-specific evidence base is still developing. No large-scale mortality study has yet been published that directly links contactless vitals data captured at application to claims outcomes. That study will come as carriers accumulate enough data from early implementations. In the meantime, the clinical evidence for the underlying biomarkers is strong, and the logical connection to underwriting fraud detection is straightforward.

The future of biometric fraud prevention in underwriting

The direction is clear even if the timeline isn't precise. As accelerated underwriting programs become the norm rather than the exception, carriers need real-time biometric verification methods that work in digital-first application flows. Lab work and nurse visits are going away. EHR and Rx data catch what's in the medical record. Contactless vitals catch what's happening right now, in the applicant's body, at the moment they apply.

The technology is also getting more capable. Early rPPG implementations measured heart rate and that was about it. Current systems capture heart rate variability, respiratory rate, blood oxygen estimation, and blood pressure indicators. As algorithms improve and training datasets grow, the precision of these measurements will increase, and the fraud signals they can detect will become more specific.

Companies like Circadify have developed contactless vital sign measurement specifically for integration into insurance underwriting workflows. The 60-second smartphone scan fits into existing digital application flows and produces real-time biometric data that complements the backward-looking data from EHR, Rx, and claims sources.

Frequently asked questions

Can an applicant cheat a contactless vitals scan?

It's much harder than lying on a questionnaire. The scan reads involuntary physiological signals — blood flow, heart rhythm, respiratory patterns — that aren't under conscious control. Presentation attacks (using a photo or video instead of a real face) are detectable through rPPG liveness verification, since fake images don't produce genuine blood flow signals. No system is foolproof, but the barrier to fraud is significantly higher than self-reported data.

Does a contactless scan replace traditional underwriting evidence?

No. It adds a data layer that didn't exist before in digital underwriting. It works alongside EHR, prescription histories, claims data, and MIB checks. Think of it as restoring part of the real-time health verification that paramedical exams used to provide, without the cost and delay.

How accurate are camera-based vital sign measurements?

Accuracy varies by implementation and measurement type. Heart rate measurement via rPPG is mature and well-validated across dozens of clinical studies. Blood pressure estimation is newer and less precise but improving. For underwriting purposes, the measurements don't need clinical-grade precision to have value — they need to be accurate enough to flag discrepancies that warrant additional investigation.

What about privacy concerns with facial video capture?

The camera scan captures a short video segment that's processed to extract vital sign data. Implementations can be designed so that the raw video is processed on-device and only the derived health metrics are transmitted to the insurer. Privacy frameworks like BIPA (Biometric Information Privacy Act) and state-level regulations set requirements for consent, data retention, and use limitations that carriers must comply with.