In the era of the digital economy, algorithms have become a tool for shaping social power structures. Masking their inherent value biases behind a fa?ade of technological rationality, they exert an invisible yet systematic form of control through their externalization as social rules, panopticon-style surveillance and autonomous decision-making. However, digital subjects do not remain entirely compliant; instead, they adopt resistance strategies such as active resistance, passive defense and organized protest, creating a dynamic struggle between technology and users. This process exposes deep-seated contradictions between the core architecture of social regulation and algorithmic rules, the protection of users' personal information and algorithmic panopticon surveillance, and users' right to autonomous choice and algorithmic contextual guidance. To address the challenges of governing algorithmic power, it is necessary to understand the core operational mechanisms within algorithms. By optimizing algorithmic design as a foundation, we can implement source-level governance to weaken algorithmic discipline; by limiting surveillance scenarios and the scope of data collection, and by establishing digital tracking notifications, we can impose core constraints on algorithmic surveillance; and by utilizing internal and external collaboration to decentralize power, we can weaken the autonomy of algorithmic decision-making. Ultimately, this will achieve the adaptation of algorithmic power rules, the safeguarding of users' fundamental rights, and the healthy and orderly development of the platform economy.