Development of systems with an evolving level of autonomy is among the most controversial and yet promising aspects of military forces’ modernisation. The application of quasi- artificial intelligence in combat machines opens a new uncharted area of possibilities. In essence, these modifications aim to reduce the extent of human interference in the functioning of unmanned systems virtually to none. The potential risks are explored in dangerous scenarios, which consider that apart from off ering far-reaching advantages of using platforms with autonomic capabilities, they can be used against humanity. These scenarios assume a pivotal role in forecasting possible directions for development of the armed forces. This paper attempts to determine the essence of combat systems autonomy while focusing on a few of the most sensitive issues. For a transparent and credible debate on combat systems autonomy, it is adviseable that no ambiguity is present in terminological, ethical, legal and technological complexities, whose misinterpretation may become a source of unnecessary understatements or even lead to distortion of the debate.