Launched in July 2024, Death Clock is an application that uses artificial intelligence (AI) to predict when its users will die. Death Clock trained its AI model using over 1,200 life expectancy studies. It then uses the answers from a questionnaire about the user’s physical health, like diet and exercise, to calculate each user’s date of death. While users of the free version will only receive this date, users of the paid version will receive lifestyle recommendations to help them live longer.

Although the AI model includes a large amount of data, the data collected from individual users is currently limited. The current questionnaire is brief and does not delve extensively into family history or lifestyle habits. Including this additional data is likely necessary to receive more accurate results from the model. Improving the model’s accuracy is key for the economic calculations of different organizations, like the government and insurance companies. For example, people can better determine if they have saved enough for retirement.

However, the increase in data collection comes at a risk — namely, user privacy and discrimination. Collecting more data for analysis and inclusion in the model exposes the data to a greater likelihood of being leaked if proper security and storage procedures are not followed. Additionally, implicit biases in the model may produce harmful outcomes (e.g., higher insurance premiums) for certain consumer groups. Therefore, it is crucial that models are developed with a diverse group of stakeholders and are used in a fair, unbiased, and privacy-conscious way.