Roger Mallett [Staff Author].
A controversial deal between tech giant Google and the National Health Service (NHS) will allow artificial intelligence units access to 1.6 million confidential medical records. Since 2014, Google has partnered with several scientists in an attempt to understand human health, but a new report reveals the data gathering goes far beyond what was originally anticipated.
According to documents obtained by the New Scientist, the data sharing agreement between Google-owned artificial intelligence company DeepMind and the Royal Free NHS Trust gives access to the sensitive healthcare data of millions of NHS patients. The chilling and wide-reaching deal allows DeepMind access to the medical records of the 1.6 million people passing annually through the three London hospitals owned by the Trust — Barnet, Chase Farm, and the Royal Free.
The Google-owned A.I. firm announced in February it was working with the NHS to build an app called Streams — intended to help hospitals monitor patients with kidney disease. However, the new information has revealed that the extent of the data being shared goes much further and includes logs of day-to-day hospital activity, records of the location and status of patients, and even logs of who visits them and when.
Results of pathology and radiology tests are also shared, as is information from critical care and accident and emergency departments. In addition, DeepMind’s access to the centralised records of all NHS hospital treatments in the U.K. means the tech company can access historical data from the last five years, all while receiving a continuous stream of new data.
At the same time, DeepMind is developing a platform called Patient Rescue, which uses hospital data streams to build tools to carry out analysis and support diagnostic decisions. The New Scientist explained how it works:
Comparing a new patient’s information with millions of other cases, Patient Rescue might be able to predict that they are in the early stages of a disease that has not yet become symptomatic, for example. Doctors could then run tests to see if the prediction is correct.