The expanding presence of algorithmic systems across governance and everyday social practices was shown to indicate a profound transformation of contemporary power relations and the rationality of governance in digital societies. The aim of the article was to conduct a philosophical analysis of the transition from disciplinary power to algorithmic power and to clarify how this shift redefined the conditions of subject formation in the twenty-first century. The study was based on a combination of hermeneutic interpretation, critical theory, and conceptual analysis, which were applied to reconstruct disciplinary models of power, examine transitional forms of control, and analyse contemporary regimes of algorithmic governance as socio-technical configurations. It was established that algorithmic power functioned through automated decision-making and predictive modelling, thereby transforming the epistemic foundations of governance and rearticulating the relationship between power, knowledge, and subjectivity. It was demonstrated that, unlike disciplinary power relying on institutional visibility and normalisation, algorithmic power operated in a decentralised and opaque manner and was embedded in digital infrastructures mediating everyday practices. It was analysed that algorithmic governance introduced an anticipatory temporal logic, whereby future behaviour rather than past action became the primary object of regulation. It was also shown that responsibility became diffused across distributed socio-technical systems, complicating classical ethical and political models. The results of the study could be applied in social philosophy, digital governance research, and the ethics of artificial intelligence to support the development of normative approaches to transparency and possibilities of critique and resistance in algorithmically mediated societies