Dilemmas and Path Optimization of Algorithm Governance in the Platform Economy: A Study Based on Stakeholder Theory

Abstract

Algorithmic systems have become the core infrastructure of platform economies, shaping labor allocation, pricing, content distribution, and risk control. While algorithms enable efficiency and scalability, they also generate persistent governance dilemmasopacity,
accountability gaps, discrimination risks, labor intensification, and fragmented regulatory oversight across jurisdictions. This qualitative empirical study uses stakeholder theory to diagnose why algorithm governance repeatedly "fails in practice" even when transparency or compliance mechanisms exist. Drawing on comparative document analysis of authoritative regulatory and standards-based instrumentssuch as the
EU Digital Services Act (DSA), the EU AI Act, the EU Platform Work Directive, China's algorithm recommendation rules, New York City's
AEDT requirements, Canada's Algorithmic Impact Assessment, and risk-management frameworks (NIST AI RMF, ISO/IEC 23894)the
study codes governance obligations and maps them to stakeholder salience (power, legitimacy, urgency). Findings indicate that platform firms'
structural power (data control, model secrecy, and design authority) interacts with regulators' information asymmetry and users/workers' weak
bargaining positions, producing three recurring dilemmas: (1) transparency paradox (disclosure without intelligibility), (2) accountability fragmentation (diffuse responsibility across platform, vendor, and deployer), and (3) compliance-performance gaps (formal compliance without
measurable harm reduction).