Artificial intelligence is challenging frameworks for both ethics and human rights and becoming a hot topic for both societal and scientific debate. AI is also placed at the top of the agenda for research and innovation policy makers. A recent publication in Technological Forecasting and Social Change presents the results from a Delphi study, inviting 231 AI experts to share their views. 

We know that there are privacy concerns related to artificial intelligence, and risks related to bias and discrimination, safety and security, economic distribution, democracy and warfare. There have been a number of suggestions for ways to address those issues. The Delphi study contributed to the identification of the most pressing issues, and prioritising among the plethora of proposed mitigation strategies. 

According to Bernd Stahl, Human Brain Project Ethics Director, the findings suggest that we need to rethink the ethics narrative around AI, which is currently focusing on specific issues. The results demonstrate the difficulty of defining clear priorities. The most popular mitigation measures coming out of the three-round Delphi study were broad and generic. None of the legislative or regulatory measures that have been proposed were rated highly. 

The Delphi study was conducted together with researchers in the SHERPA project, a partnering project of the Human Brain Project, working specifically on questions of ethics of AI. With some of the underpinnings of the Delphi study inspired by earlier work on AI and ethics form the HBP, including an assessment of the ethical and social concerns of artificial intelligence in neuroinformatics research. In the paper, the authors conclude that the debate around ethics and human rights of AI would benefit from such a reframing, with a stronger emphasis of the systems nature artificial intelligence ecosystems.

Want to read the paper in full? 

Stahl BC, Brooks L, Hatzakis T, Santiago N, Wright D, Exploring ethics and human rights in artificial intelligence – A Delphi study, Technological Forecasting and Social Change, 2023;191. DOI: 10.1016/j.techfore.2023.122502

You can also read more about this work on the University of Nottingham’s Responsible Digital Futures blog: What Do Experts Think about the Ethics of AI?