Dr Diana Martin - SEFI Ethics SIG, TU Eindhoven Dr Diana Martin (co-chair SEFI Ethics…
Dympna O’Sullivan, Damian Gordon, Ioannis Stavrakakis (TU Dublin)
Project Website: http://ethics4eu.eu/, Twitter: @Ethics4EU
If you applied for a loan, would you be happy to learn that the decision had been taken by a computer? If you underwent a medical test, would you be satisfied to learn that the diagnosis was made by a computer? If you were found guilty of a crime, would you be assured to learn that your sentence was going to be determined by a computer?If you think these three questions sound like fanciful scenarios, there are not, and they are already happening in some countries around the world on a daily basis (Sachan, et al., 2020; Sounderajah, et al., 2021.; Deeks, 2019); where a wide range of organisations are becoming increasingly reliant on computer systems to take decisions for them. This reliance on technology inevitably leads to a wide range of ethical concerns and considerations, and not just in the field of artificial intelligence systems, but in other areas such as: the seemingly irreparable issues of bias in facial recognition software (Libby and Ehrenfeld, 2021); the collection, sale and misuse of private data (Farooqi, et al., 2020); the problems of social media with fake news and cyberharassment (Rafee, 2020, p. 173 in book); the never-ending dataleaks from cloud storage (Jartelius, 2020); and the decisions that autonomous vehicles may have to make as to who they will save and who will they let die (Bigman, 2020).
These types of computer ethics issues have become so serious that a number of third-level institutes across Europe are collaborating to explore some of these key computer ethics dilemmas, and to develop educational content that is based on pedagogically sound principles, to ethically education the next generation of software developers, as part of the Erasmus+ Ethics4EU project (O’Sullivan and Gordon, 2020). One aspect of the project was to undertake a survey in 61 universities, across 23 European countries, to ask their Computer Science departments a number of questions on computer ethics (including whether or not they teach ethics on their computer programmes, and whether or not they think it is an important topic). The key findings of the survey were:
- Over one third (22 out of 61) of those institutions do not teach computer ethics on any of their programmes, even though the vast majority of them agreed on the importance of the topic itself.
- Where computer ethics is not taught as part of Computer Science (or related) programmes, the most common reasons cited are either a lack of staff availability, or a lack of expertise in the domain.
- Computer ethics is considered more important for some computer science subjects than for others, and the most commonly cited topics that ethics should be taught in were: Data Science, Artificial Intelligence and Computer Security.
- Most institutions devote a relatively small number of hours to teaching computer ethics on their Computer Science or related programmes, 67% of institutions surveyed teach 10 hours or less per semester.
- In a significant number of responses where computer ethics was being taught, it is taught as a standalone subject, and not integrated throughout the curriculum.
More detailed results can be found in our study Stavrakakis, et al. (2021), where it is clear that these findings demonstrate a strong need for the development of an open set of computer ethics teaching materials that combine the expertise from several Computer Science departments, but also in partnership with Ethics departments and other related disciplines, to bring many voices together on this vitally important topic.