Google’s DeepMind acknowledges it ‘fell short’ in ‘illegal’ NHS deal
A controversial partnership between the NHS and Google’s AI development lab DeepMind has been branded ‘illegal’ by The Information Commission (ICO) after it found that insufficient safeguards had been put in place to protect confidential medical records.
Google’s DeepMind acknowledges it ‘fell short’ in ‘illegal’ NHS deal
The Royal Free NHS Foundation Trust embarked on a tie-up with DeepMind by agreeing to share 1.6m patient data records with the firm in the hope of training an algorithm to better identify patients who may be at risk of going on to develop kidney disease.
This partnership was rebuked by an earlier independent review which found that the data-sharing agreement contained ‘deficiencies’, prompting a revision to the text which has itself now been criticised as being in need of ‘a lot of changes’.
In its latest salvo of criticism the ICO reserved much of its criticism for the Royal Free but Dr Julian Huppert, chair of the DeepMind Health Independent Review Panel, also had a warning for DeepMind itself, arguing that it had a duty to ‘remind partners of their responsibilities’.
Huppert stressed that DeepMind hadn’t fallen foul of any laws but said: “People are concerned about the power of big technology firms, and we felt that we should hold DeepMind to a very high standard because of its link to Google.”
In a blogpost DeepMind co-founder Mustafa Suleyman wrote: "There's a fine line between finding exciting new ways to improve care, and moving ahead of patients' expectations. We know that we fell short at this when our work in health began, and we'll keep listening and learning about how to get better at this.”
Amongst the issues identified by the panel was a flaw in the data centre enabling anyone with access to overwrite data and potentially upload malware while devices used by doctors to access patient records lacked the most up-to-date security measures.
Google recently doubled the number of healthcare staff on its DeepMind programme in recognition of the potential growth in the field.