Library Element Report

Ethical, Social, and Political Challenges of Artificial Intelligence in Health

Uploaded by RRI Tools on 10 May 2018

Ethical, Social, and Political Challenges of Artificial Intelligence in Health. A report by Future Advocacy. Written and researched by Matthew Fenech, Nika Strukelj, Olly Buston for the Wellcome Trust, April 2018

AI is coming to healthcare and medical research. How do these technologies challenge long-standing principles of medical ethics? What new challenges will healthcare practitioners, medical researchers, patients, and regulators need to face?

We interviewed and organised roundtable discussions with more than 70 experts all round the world, including clinicians, patients, computer scientists, policymakers, ethicists and regulators. 

We identify 5 key use cases for AI in health and medical research, separating hype and hope from what's actually happening:

  • Process optimisation     e.g    procurement, logistics, and staff scheduling
  • Preclinical research  e.g   drug   discovery and genomic science
  • Clinical pathways    e.g.   diagnostics and prognostication
  • Patient-facing  applications  e.g delivery of  therapies or the provision of information
  • Population-level applications  e.g. identifying epidemics and understanding non-communicable chronic diseases 

We also identify 10 ethical, social, and political challenges that urgently need to be addressed. It is essential that research focusing on these challenges is multidisciplinary, and that the the voices of patients and their relatives are heard. It is only by developing tools that address real-world medical needs that the opportunities of artificial intelligence in health can be maximised, while the risks are minimised.

Across these use cases, a number of ethical, social, and political challenges are raised and the 10 most important are: 

  1. What effect will AI have on human relationships in health and care? 
  2. How is the use, storage and sharing of medical data impacted by AI? 
  3. What are the implications of issues around algorithmic transparency/explainability on health? 
  4. Will these technologies help eradicate or exacerbate existing health inequalities?
  5. What is the difference between an algorithmic decision and a human decision? 
  6. What do patients and members of the public want from AI and related technologies? 
  7. How should these technologies be regulated?  
  8. Just because these technologies could enable access to new information, should we always use it? 
  9. What makes algorithms, and the entities that create them, trustworthy?
  10. What are the implications of collaboration between public and private sector organisations in the development of these tools?       

In  this  report,  we  explore  these  and  the  other challenges  raised  by  our  research  and  make recommendations   for   further   study   in   this  complex  and  sensitive  field.  We  also  find  that there  are  overarching  ethical  themes,  namely  consent, fairness and rights, that cut across the challenges  we  identify.  We  ask  how  users  can give  meaningful  consent  to  an  AI  where  there may be an element of autonomy in the algorithm’s decisions, or where we do not fully understand these   decisions.   Ensuring  fairness   through preventing  and  eliminating  health  inequality, and  providing  value  to  stakeholders  is  another critical issue. Finally, the right to health may well be  expanded  to  encompass  questions  such  as “do people have a right to know how much AI is used in their care?” and “do people have a right not to have AI involved in their care at all?”

We recommend a multidisciplinary approach to dealing with these issues. This refers not only to galvanising a broad range of experts, many of  whom will use and be impacted by these tools, but also to the active participation of patients, their relatives, and the public in their development. It is equally important that these technologies are developed with a view to sharing their benefits as  widely  as  possible.  This  is  the  best  way  to ensure that real-world challenges are addressed, that  the  needs  of  patients  beyond  their  clinical care are considered, and that these technologies are accepted by patients and practitioners alike.








Related Resources