Google has been in the hot seat this week over allegations that it was using millions of people’s health data to feed and build its own artificial intelligence (AI) health tools.
The news, reported in The Wall Street Journal on Monday (Nov. 11), prompted an inquiry by the U.S. Department of Health and Human Services (HHS) Office for Civil Rights.
Now, Bloomberg is reporting that Google is pushing back against the allegations.
David Feinberg, head of Google Health, said employees of the company only have patient information to help build a new search tool for the Ascension hospital network, and that no patient data is being used to aid in Google’s AI research.
Feinberg said its contract with Ascension is operating under the constraints of U.S. health privacy law, and that it has access to the files for the sole reason of organizing health records systems to facilitate easier searching.
“That’s all we’re allowed to do and that’s all we are doing,” he said.
The HHS Office for Civil Rights said it “would like to learn more information about this mass collection of individuals’ medical records with respect to the implications for patient privacy under HIPAA.”
HIPAA is the Health Insurance Portability and Accountability Act, which guarantees confidentiality of health records and governs how and when they can be shared.
Google said only Ascension employees have access to the health data.
“All data is logically siloed to Ascension and housed within a virtual private space encrypted with dedicated keys,” said Thomas Kurian, chief executive officer of Google Cloud. “Google does not sell, share or otherwise combine data from Ascension with any other data.”
Senator Richard Blumenthal of Connecticut called Google’s activity a “blatant disregard for privacy” as well as “beyond shameful.”
Kurian said that’s not the case.
“We never actually have Google employees understand individual patients’ data when it goes into the model. We have other technologies that de-identify it,” he said.