Legal Priorities for the Development of Artificial Intelligence and Their Implementation in the Sphere of Education

The paper examines the conceptual foundations of the legal regulation of artificial intelligence and the possibilities of their application to the education system. The initial premise of the work is the proposition that AI has ceased to be a private technical instrument and has turned into an infrastructural technology that defines a new socio-political era and generates constitutional challenges of a systemic, synergetic and existential nature. The author identifies the key threats: the substitution of democratic institutions by algorithmic governance, technologically conditioned inequality and algorithmic discrimination, micro-manipulation of the information flow, and the concentration of power resources among the holders of algorithms. As a meta-threat, the challenge to the socio-humanistic paradigm is considered, associated with the choice between transhumanism and classical humanism. By way of discussion, eight principles of the legal regulation of AI are formulated: state strategic guidance and technological integrity; guaranteed protection of the affected person and enhanced liability; "human in the loop" in critical decisions; anthropological priority; digital equality and the prohibition of discrimination; respect for dignity and non-interference in the private sphere; managed transparency; technological neutrality and anticipatory regulation. Further, these principles are projected onto the sphere of education. The author distinguishes two modalities of the use of AI — "a simulator for the mind" and "a prosthesis for laziness" — and shows that legal regulation must establish the boundaries of the permissible application of the technology. The legal risks of algorithmic assessment are analyzed, along with the problem of the right to explanation and the protection of the student from unfounded accusations, digital inequality in access to AI tools, threats to privacy and dignity, as well as the question of the status of the educator as a condition for the humanistic development of AI. A conclusion is drawn regarding the necessity of building a system of requirements for transparency, accountability, equality of access and the preservation of human control in critical decisions.