Abstract: The widespread application of artificial intelligence in administrative sectors has aroused public concerns regarding transparency,fairness,and accountability.External law directly addresses algorithmic challenges to administrative operations and seeks to mitigate risks stemming from digital systems through legislative intervention or judicial review.Nevertheless,it fails to effectively monitor and respond to the hybrid integration of algorithms and administrative practice,which often renders accountability futile.Where external oversight of digital administration proves ineffective,exploring internal administrative transformations—particularly the formation and normalization of regulatory structures embedding AI algorithms into public administration—carries profound theoretical and practical implications.Internal administrative law regards AI algorithms not merely as discrete regulatory objects,but analyzes them as integrated administrative structures.It prioritizes cyclical supervision over the design,integration and assessment of digital systems to fulfill the normative requirements of democratic rule of law.Furthermore,it emphasizes coordination between internal and external legal regimes,calls for the gradual development of digital administrative law,and promotes targeted reform and innovation of administrative law.

Key words: digital administration, artificial intelligence, internal administrative law, regulatory structure, algorithmic impact assessment