Unlike existing studies on labour and income in the digital era, this paper argues not only that the impact of the digital economy’s intervention in the labour process is fragmented rather than comprehensive, but also that the transformation of job demand and labour supply behaviours is simultaneous and related to the attributes of the industries in which they operate. Drawing on the conventional biased technological progress hypothesis and labour process theory, we argue that the digital economy has generally increased the labour income inequality for migrant workers in China. Using geospatially matched China Labour Dynamics Survey 2018 microdata and provincial digitalisation indices, we uncover a digital ‘upgrading trap’: the development of the digital economy hides the process of inequality formation in the hedging relationship between objective labour demand ‘upgrading’ and subjective labour supply ‘expanding’. The former can be summarised as the risk of ‘no job’ and the latter as the risk of ‘no way back’. Counterintuitively, consumer Internet development demonstrates a greater role in both reducing workers’ inequality in secondary labour markets and promoting a fair primary distribution. These findings reconceptualise digital inequality as coevolutionary outcomes, and offer a tripartite governance way for inclusive growth through portable skill certification, algorithmic accountability mechanisms, and interoperable social security systems.