The unpaid auditors: Trust and De-influencing in E-Commerce

Abstract

In digital commerce, AI-driven recommendation systems increasingly decide which products, influencers, and stories a consumer sees, and that visibility carries a trust problem of its own. When consumers sense saturation, staged authenticity, or hidden commercial motives, suspicion sets in and they begin checking claims themselves, scrutinizing details, flagging what does not add up, warning others, and pointing to alternatives. Existing work usually discusses de-influencing, digital labor, and e-commerce adoption separately, leaving the unpaid evaluative work that consumers perform around influencer content largely untheorized. We address that gap by naming this unpaid checking audience correction labor and arguing that it has become one of the ways trust now gets built in algorithmic e-commerce. The construct works at two levels. For the individual, it appears as consumer due diligence, where influencer content is treated as evidence to be tested rather than advice to be followed. For the market, it becomes reverse influence, as audience reactions travel back and reshape how credible influencers, brands, products, and platform recommendations appear. The article makes three contributions, an operational definition of audience correction labor and its two dimensions, a four-layer framework linking AI-mediated visibility, trust conditions, correction labor, and trust recalibration to e-commerce adoption, and a set of eleven testable propositions. Read this way, de-influencing looks less like anti-consumption and more like trust-building work, a shift that changes what the behavior means in practice. We close with the Moroccan case, where checking that happens off-platform, in WhatsApp and Facebook groups, weighs heavily on online purchasing because trust in formal institutions remains low. AI does not only make e-commerce more convenient; it also raises credibility problems that consumers resolve by correcting one another, and this matters most where trust is the barrier to buying online.

Classification JEL: M31, M37, D12, L81.

Paper type: Theoretical Research.