To be able to understand the possible implications of a new technology one must have tools to do the analysis. While a good question to consider is always, "who benefits and who pays", it is not sufficient on its own. Some work has been done on the content issue by Project ImpactCS[3] on content, but I would argue that it is useful to present specific models for analysis that allow students to derive the more specific "ethical issues". In the course that I teach I discuss models of analysis from economics, sociology, law and philosophical ethics.
Cost benefit analysis and utilitarianism are common tools used to analyze design decisions. While often useful, there is the occasional flaw with basing all of ones' decisions on utilitarianism. The Ford Pinto case is an example of cost benefit analysis being misused.
To balance utilitarian ethics, I also discuss Kant's categorical imperative and his concept of respect for persons. The first is useful in conveying the idea that one should think about how the proposed technology would impact on the designer if the designer had to be the user. The concept of respect for persons, upon which informed consent is based, is necessary for anyone in a professional role.
Several models of distributive justice are also described in the course. The purpose of these models is to reinforce the idea of designing a technology as though you would be required to use it, as well as to develop an understanding that a new technology may impact on (apparently) completely unrelated people.
The reference to sociology is to ensure that people consider the impact of work related technologies on the skill levels of the workers. Even highly skilled people can suffer from the implementation of new technologies[1] or the implementation of very "prescriptive" technologies[2]. The alienating nature of prescriptive technologies, as well as challenges to the question of whether informed consent is informed consent are covered by basic sociological models, as are organizational structure and function.
Perhaps the most important issues for the professional life of a computer scientist or an engineer are whistle blowing, liability issues, professional duty, and intellectual property. Well reasoned views on these topics, topics that are some of the "ethical issues" described in [3], all can be derived from the discussion of the previous models. Whistle blowing out of respect for persons, utilitarianism (depending on how one calculates utility) and Marx's conflict theory. The discussion of liability naturally follows from the distributive justice models as well as utilitarianism. Professional responsibility and duty to ones client can be easily derived from all of the ethical models and intellectual property follows from utilitarianism and Kant's categorical imperative which is a simple enough concept, but one which many people do not follow. I find myself constantly asking computer science students the following question when they ask me for a copy of some piece of software, or they come to me with a photocopied text book: "you plan to write computer programs for a living don't you? How good of a living would you have if one person bought your program and the rest copied it for free?"