How can technological innovation be designed in such a way that ethical responsibility is not neglected? The bidt project “Ethics in Agile Software Development” (EDAP) has not only raised awareness of this issue but also developed concrete solutions. Rather than viewing ethics as a post-hoc review or a moral add-on, EDAP has brought it into the very place where decisions about digital systems are actually made: the development process itself. The focus was on embedding ethical reflection methodologically in such a way that it becomes relevant to action even under the conditions of agile software development.
With “Introduction to Ethical Software Development”, a book by the project team is now available to mark the conclusion of the project. In it, Dr Jan Gogoll and Dr Niina Zuber summarise the project’s key findings in a concise, freely accessible format. Published by Springer, this open-access volume demonstrates how the interface between technology and ethics can be conceptualised from the perspective of development – that is, from the perspective in which decisions regarding the goals, priorities and impacts of software take concrete form. The following video interview takes these considerations further: in it, Gogoll and Zuber discuss responsibility in day-to-day development, the organisational prerequisites for ethical deliberation, and the question of why human judgement remains indispensable, particularly in the age of AI-generated code.
By loading the video, you agree to YouTube’s privacy policy.
Learn more
What organisational measures are needed to ensure that ethical considerations can genuinely take place in the day-to-day work of software developers?
Jan Gogoll:
For ethical considerations to actually take effect in the day-to-day work of software developers, more is needed than good will or general appeals. The organisational framework is crucial: in other words, time, resources and incentives to be able to raise and address ethical questions in the first place. If development processes focus exclusively on short-term goals such as efficiency or ‘time-to-market’, ethical considerations have a hard time gaining traction.
Ethical software development therefore requires structures in which such questions are raised at an early stage and taken seriously – for example, during planning, through clearly defined roles and within an open team culture. Organisations often already possess ethical knowledge; the crucial thing is that this knowledge is actually put to use. Ethical software development should therefore not be the heroic feat of individuals, but rather be made possible at an organisational level and embedded within the process.
Why is ethical software development so important today?
Niina Zuber:
Ethical software development is so important today because software shapes almost every aspect of our daily lives. It influences how we work, communicate, learn and make decisions. That is why software should not only be technically capable, but also fair, secure, transparent and accountable.
It is not just a matter of avoiding harm. Well-developed software can also make a very concrete positive contribution: it can support people, promote participation, strengthen trust and improve everyday life. Ethical software development therefore means not only limiting risks, but consciously designing software that is socially beneficial.
AI coding assistants such as Claude Code or Codex now write code themselves – often faster than humans. Does this change the ethical responsibility of developers? Or does it actually make it even more important?
Jan Gogoll:
AI coding assistants make ethical responsibility more important rather than less so. Whilst they can generate code very quickly, they do not understand the social or moral context in which this code will later take effect. They make suggestions, but do not take responsibility – neither for data protection nor for fairness, security or potential harm.
The task of making normative judgements therefore remains with humans. The role of developers is thus changing: away from simply writing every single line of code, towards greater scrutiny, selection and responsibility. Precisely because AI accelerates development, there is an even greater need for human judgement – so that development is not just faster, but also accountable.
To the Research Project
To The Open-Access-Book
