This article has been reviewed according to Science X's editorial process and policies. Editors have highlighted the following attributes while ensuring the content's credibility: AI engineers often recognize ethical risks associated with the systems they build, but many lack the authority, incentives and organizational support needed to act on them, according to new research from the University of Manchester. Based on in-depth interviews with AI and software engineers working across technology, finance, semiconductor manufacturing and research organizations, the study found that engineers could readily identify issues such as inaccurate outputs, unfair automated decisions, AI systems that are difficult to explain, and the use of automated judgment in areas that can significantly affect people's lives.
However, many also described safeguards they would like to implement but felt unable to put into practice. What they lacked was not awareness. It was the structural capacity to act on it.
The researchers describe this as "ethical awareness without ethical agency": engineers who know the right thing to do but are unable to act on it. Presented at the 10th Data for Policy Conference, the research raises questions about whether current oversight frameworks measure genuine ethical practice or merely the documentation surrounding it. The study uncovered a series of organizational factors limiting engineers' ability to act, including box-checking compliance processes, commercial and deadline pressures, and reward structures that prioritize speed over rigor.
Together, these create what the paper calls "compliance theater": organizations that signal ethical commitment without consistently putting it into practice. "Most people assume that AI ethics is a problem of what engineers know or care about. However, the people I interviewed care and they know.
What they described was the experience of working inside organizations where raising concerns costs you professionally, and where doing the work that ethics actually requires is not what gets rewarded. "If you want AI to be built ethically, you have to change the conditions under which it gets built. Training the individual engineer harder is not going to make the difference," said Alessia Vlasceanu, the study's lead researcher based in the Department of Computer Science, the University of Manchester.
At a time when governments and organizations are introducing new AI rules and standards, including under the EU AI Act, the study suggests that many current efforts to govern AI focus on producing documents, policies and reports that demonstrate ethical commitment. However, the findings indicate that unless organizations also change how AI is developed in practice, these measures may amount to little more than a box-checking exercise. Professor Caroline Jay, who supervised the research, said, "This is not a story about bad companies or bad engineers.
Extract — continue reading at the source.
