Ethical AI - Perspectives from Patient Advocates: Ethics and Emerging Technology – Group 4

This is the final blog in a series of 5 which discusses the results of the Ethical AI workshop at MPNE Consensus Data 2024.

The workshop was held in collaboration with members of the iToBoS project and a group of patient advocates from the wider MPNE community, drawn from across Europe.

The patient advocates offered a wealth of subject matter expertise on ethical topics such as governance, accountability, privacy, fairness, trust and transparency. After the participants were divided into 4 groups, they were provided with a cartoon, and asked to discuss their experiences and perspectives based on the topic or theme depicted.

This blog focuses on the discussion from Group 4 based on the cartoon below. 

Using Slido, the following 12 responses from 4 participants were collated when the participants were asked to speak openly, freely and honestly about the cartoon. In this instance, one participant acted as the moderator for the group and entered key words on others behalf.

This picture depicts a “Robo prison: Inside a Faraday Cage the machine would be cut off from Internet signals and helpless – but do really want that?” The author was trying to explain how it’s important for the public to see how beneficial AI technology is for drug discovery or medical treatments[1]. “Oppenheimer” suggests the participants may view AI technology as a benefit to humans in the present, but like “Oppenheimer” and the atomic bomb, it’s “Unknowns” may have more sinister repercussions in the future.

Similarly, language such as “Power”, “Risk”, “Tradeoff”, and “Ambiguity” all carry negative undertones suggesting the participants maybe sceptical as to what, and how, technology will be implemented, secured, or controlled.

However, on the other hand, concepts such as “Hope”, “Progress” and “Utilitarian” evoke more positive vibes towards AI and its potential benefits. The participants may view it as a “Pivotal” event in an era of emerging technologies. If managed correctly, and in the right “Context”, it may elevate and change healthcare in the future, for the greater good.

During this blog series, “Trust” has been key theme within each groups’ discussion. This highlights the importance of “Trust”, and what it means to the public and in the provision of healthcare. All three previous groups alluded to the participants’ lack of “Trust” in data quality, data storage, data security, accessibility rights, and how data will be shared. Similarly Group 4 has used language such as “Power” and “Tradeoff” inferring it can be used as leverage and potential harnessed inappropriately by bad actors.

Taking into consideration all four groups’ discussions and their combined subject matter expertise, iToBoS should be diligent in showing transparency, accountability, efficiency, explainability and interpretability, accuracy, and fairness in the iToBoS solutions to the public.

In addition, iToBoS should continue with a governance board whose responsibility it is to ensure robust policies and standards of practice are implemented. If these recommendations from the patient advocacy group are incorporated into iToBoS as the technology matures, it may help alleviate some of the fears surrounding the use of AI in healthcare. It may also encourage the development of “Trust”, which may help with engagement and overarching uptake of the iToBoS solutions with the public.

[1] Iyad Rahwan, ‘© Iyad Rahwan, EvilAICartoons.Com’, n.d., https://dldnews.com/ai-cartoons-iyad-rahwan-interview/.

More about Ethical AI workshop at MPNE Consensus Data 2024 here