Insurance Companies Know When You’ll Die – They Use Machine Learning to Predict Your Expiration

Danish AI model analyzes 6 million life histories to predict death with 78% accuracy, raising privacy concerns

Annemarije de Boer Avatar
Annemarije de Boer Avatar

By

Our editorial process is built on human expertise, ensuring that every article is reliable and trustworthy. AI helps us shape our content to be as accurate and engaging as possible.
Learn more about our commitment to integrity in our Code of Ethics.

Image credit: Wikimedia

Key Takeaways

Key Takeaways

  • Danish AI predicts four-year mortality with 78% accuracy using 6 million life histories
  • Insurers deploy smartwatch data to contest claims during two-year contestability periods
  • Mortality algorithms embed zip code and income biases into coverage decisions

Your fitness tracker measures more than steps and sleep cycles. Hidden in that heart rate data lies something far more valuable to life insurance companies: a prediction of when you’ll die. Danish researchers built an AI system called life2vec that analyzed 6 million people’s complete life histories and predicted four-year mortality with 78% accuracy—outperforming traditional insurance models by 11%. The unsettling part? Insurers are racing to deploy similar tools using your everyday digital exhaust.

The Algorithm Behind the Curtain

Life insurance companies are quietly replacing human underwriters with AI death calculators.

The life2vec model reads like science fiction made real. It transforms everything from medical visits to job changes into sequence data, then predicts survival with transformer architecture—the same technology powering ChatGPT. Society of Actuaries research shows insurers are developing AI models that ingest “non-traditional data” beyond typical medical exams, including detailed treatment records and lifestyle indicators. Your Apple Watch’s irregular heartbeat alerts are prime input for mortality scoring algorithms.

From Data Streams to Denial Letters

Wearable devices and health apps are feeding insurance companies ammunition for claim disputes.

Life insurance attorneys warn that carriers might soon weaponize AI predictions during contestability periods. Let’s say that you apply for coverage while seemingly healthy, then die two years later from an undiagnosed condition. The insurer’s AI, having analyzed your smartwatch data, claims their algorithm detected early warning signs you “should have disclosed.” Legal experts predict disputes where companies use third-party mortality scores to justify application denials or policy cancellations—decisions based on patterns invisible to applicants themselves.

The Bias Buried in Code

AI mortality models risk encoding social inequalities as mathematical certainties.

Sune Lehman, who led the life2vec research, explicitly warns against insurance applications, arguing they would “threaten the basic principle of shared risk.” The concern runs deeper than privacy. These models might learn that certain zip codes, job types, or income levels correlate with earlier death—then embed those biases into coverage decisions. Unlike traditional actuarial tables, AI systems operate as black boxes, making their reasoning nearly impossible to challenge or audit.

Your Digital Death Score

Every health app and fitness tracker potentially feeds an invisible mortality rating system.

While consumer “death calculator” apps offer morbid entertainment, the insurance industry’s versions carry life-altering consequences. Your Fitbit sleep scores and telehealth consultations could factor into an algorithmic assessment of your lifespan. The most disturbing aspect? You’ll never know your mortality score or how it was calculated—until coverage gets denied or claims get contested based on risks only an AI could detect.

Share this

At Gadget Review, our guides, reviews, and news are driven by thorough human expertise and use our Trust Rating system and the True Score. AI assists in refining our editorial process, ensuring that every article is engaging, clear and succinct. See how we write our content here →