Allegations against AI company DeepMind: privacy of British patients violated

A lawsuit against AI company DeepMind is being initiated in the UK. In 2015, Google subsidiary DeepMind had access to medical data from 1.6 million patients in the UK healthcare system. Now a law firm is to clarify on behalf of the patient how such sensitive data can be used in the future without violating the rights of the patient.

The allegation in the case of the data made available in 2015 is that the patients were not sufficiently informed about the process. The law firm, Mishcon de Reya, is recognized as one of the leading legal firms in the UK. BBC cited online Andrew Prismall, Principal Attorney in charge of the case: “The last thing you expect (as a patient) is for your personal medical records to fall into the hands of one of the world’s largest tech companies.” Commenting on the case, Ben Lasserson, partner at Mishcon de Reya, said: “This important lawsuit is intended to help resolve fundamental issues relating to the handling of sensitive personal data and special category data.”

DeepMind Technologies was founded in 2010 by three scientists. In 2014, the US company Google took over the British start-up; since then the AI ​​company has been operating as “Google DeepMind”. The company’s AI processes kept making headlines: In 2016, DeepMind researcher David Silver presented AlphaGo, a software that masters the extremely difficult board game Go at grandmaster level. In late September, DeepMind reported improvements in its short-term rain forecast.

As part of the cooperation with the British National Health Service (NHS), DeepMind was given access to patient data from three London hospitals. The aim of the cooperation should be to reduce the costs of medical care through automatic analyzes and to predict illnesses earlier than before.

As a first result, DeepMind 2019 presented a method to predict impending kidney failure with the help of AI. Acute kidney failure is a common complication in hospitalized patients. DeepMind researchers published in the journal Nature an automated predictive tool that should detect such problems up to 48 hours earlier than traditional methods.

In 2017, DeepMind’s cooperation with the NHS was targeted by the British Information Commission (ICO). The country’s highest data protection authority complained that the hospitals involved did not provide their patients with sufficient information about how their data had been handled within the framework of the cooperation.

When Britain’s top privacy advocate Elizabeth Denham was accused of not doing enough to protect patient privacy, DeepMind responded by saying that the NHS collaboration focused on developing medical tools, not on patient needs, reports BBC online.


To home page