The social and cultural life of artificial intelligence (AI) slop: A research framework

The proliferation of artificial intelligence (AI)-generated content across digital platforms has sparked urgent institutional responses, most focused on detection tools that treat synthetic content as a form of pollution—or more colloquially “slop”—requiring technical mitigation. We argue that such approaches are insufficient. Instead, we propose the analytical category SLOP, distinct from its vernacular dismissal, to name a condition rather than a class of objects: information ecosystems in which AI-generated content has reached sufficient scale and surface competence that the signals once used to evaluate quality no longer reliably do so. What has become scarce is not content but the means of judging it. Under this condition, the value and potential harm of any given artifact emerge not from properties innate to the content but from social context—who produces it, who engages with it, and with what motivations. Existing frameworks are useful but inadequate. Spam scholarship identifies the infrastructures of attention through which AI-generated content flows; misinformation research illuminates the participatory aspects of this content creation; and detection work can isolate the technical signals that mark AI use. Yet none accounts for the relational properties emerging around synthetic content, which is actively sought out by some users, unlike spam, and neither reliably false nor intended to deceive, as with misinformation. We characterize SLOP along three dimensions (production, reception, and effects), drawing on platform studies and cultural sociology to propose a research agenda spanning ethnographies of production, studies of reception, and analysis of institutional effects and governance.