SAN FRANCISCO: A US study published Thursday showed the privacy of people’s health data is subject to greater risks as more advances are being achieved in artificial intelligence (AI).
The study by researchers at University of California, Berkeley (UC Berkeley) said that with the help of AI tools, the status of individuals can be easily identified by monitoring their daily patterns in step data, which expose their life habits by the information gathered by activity trackers, smartwatches and smartphones in combination of relevant geographic data.
The findings of the study, which were published in the latest edition of JAMA Network Open journal, were supported by a two-year-long experiment on data mining involving more than 15,000 Americans.
The study said such privacy cannot be sufficiently protected by current laws and regulations including the US Health Insurance Portability and Accountability Act (HIPAA).
UC Berkeley engineer Anil Aswani, who led the study, said many tech companies such as Facebook can collect step data through apps on smartphones.
If Facebook matches the gathered information with health care data it buys from another company, the social media network can sell the advertising based on the health data at its hands or simply make profits by selling the private information to third parties, Aswani said.
It’s not the fault of mobile devices, but the way how the information collected by smartphones being misused, he said.
Progress in AI makes it easier for companies to gain access to people’s health data, which increases the temptation for companies to use it in illegal or unethical ways, said Aswani.
He noted that people’s health care information is an important, highly appealing data asset being pursued by employers, mortgage lenders, credit card companies and other agents who are keen on promoting classified advertising by targeting particular population groups, such as pregnant women or people with disability.
Aswani urged to improve current laws and rules like HIPAA to give full protection to people’s health data.