Search for: All records

Although development of Artificial Intelligence (AI) technologies has been underway for decades, the acceleration of AI capabilities and rapid expansion of user access in the past few years has elicited public excitement as well as alarm. Leaders in government and academia, as well as members of the public, are recognizing the critical need for the ethical production and management of AI. As a result, society is placing immense trust in engineering undergraduate and graduate programs to train future developers of AI in their ethical and public welfare responsibilities. In this paper, we investigate whether engineering master’s students believe they receive the training they need from their educational curricula to negotiate this complex ethical landscape. The goal of the broader project is to understand how engineering students become public welfare “watchdogs”; i.e., how they learn to recognize and respond to their public welfare responsibilities. As part of this project, we conducted in-depth interviews with 62 electrical and computer engineering master’s students at a large public university about their educational experiences and understanding of engineers’ professional responsibilities, including those related specifically to AI technologies. This paper asks, (1) do engineering master’s students see potential dangers of AI related to how the technologies are developed, used, or possibly misused? (2) Do they feel equipped to handle the challenges of these technologies and respond ethically when faced with difficult situations? (3) Do they hold their engineering educators accountable for training them in ethical concerns around AI? We find that although some engineering master’s students see exciting possibilities of AI, most are deeply concerned about the ethical and public welfare issues that accompany its advancement and deployment. While some students feel equipped to handle these challenges, the majority feel unprepared to manage these complex situations in their professional work. Additionally, students reported that the ethical development and application of technologies like AI is often not included in curricula or are viewed as “soft skills” that are not as important as “technical” knowledge. Although some students we interviewed shared the sense of apathy toward these topics that they see from their engineering program, most were eager to receive more training in AI ethics. These results underscore the pressing need for engineering education programs, including graduate programs, to integrate comprehensive ethics, public responsibility, and whistleblower training within their curricula to ensure that the engineers of tomorrow are well-equipped to address the novel ethical dilemmas of AI that are likely to arise in the coming years. 

Postindustrial societies are characterized by complex technological objects and systems. The publics therein are increasingly reliant on engineers to take public welfare into account when designing and maintaining these objects and systems and raise awareness when public welfare is threatened. The training engineers receive in their engineering undergraduate education is thus expected to foster their sense of responsibility to public welfare, but such training may be absent or insufficient. In this paper, we draw on a survey of 120 employed engineers in the US to assess the extent to which they received formal public responsibility training in their undergraduate education and to assess the relationships between this training and their response to one of four randomly assigned ethical dilemmas. We find that engineers who reported receiving training in public welfare responsibilities as undergraduate students felt better prepared to address public welfare issues than those who had not received such training. Individuals with training in public welfare responsibilities were less likely to identify the ethical dilemma as irrelevant to their work, indicate that such dilemmas happen all the time, be uncomfortable reporting the issue, and believe that their colleagues might respect them less if they report. These findings have implications for improving engineering ethics education and ethical conduct trainings within engineering practice more broadly.