Certain practices, such as unauthorised collaboration with other students and unreferenced copying from external sources, are generally considered in the educational context to be breaches of academic integrity. This paper explores whether there are differences between the perceptions of the acceptability of these practices in the academic context and in the professional context. From focus groups of computing academics and students, and an online survey, we find that there are indeed differences in perceptions: that many practices considered unacceptable in the academic context are considered significantly more acceptable in the professional context. This raises questions concerning the roles of summative assessment and the possibilities of authentic assessment. The paper concludes that in much of programming education there is an unbreachable rift between the goal of authentic assessment, which necessarily entails collaborative work, and the need for summative assessment of individual effort, which typically requires work in isolation. The findings of our research have implications for computing education programs, particularly in regard to preparation of students for the workforce.