Recommendation in the era of GenAI: from an attention economy to a trust economy

Abstract

The article examines the shift in the digital cultural ecosystem triggered by the widespread adoption of Generative AI. Historically, the platformization of cultural industries established an attention economy, where recommendation algorithms were essential for guiding users through overwhelming content catalogs. However, the recent influx of synthetic media has disrupted the dynamic, creating a synthetic tsunami that threatens cultural diversity, undermines user trust, and jeopardizes the livelihoods of human creators. Furthermore, this unchecked proliferation risks model degeneration for the AI systems themselves. To mitigate these economic and social tensions, Digital Service Providers are evolving their algorithmic policies. We analyze how platforms handle AI detection tools, transparent labeling, and adjusted recommendation guidelines to filter and distinguish human creations from automated content. Ultimately, this paper argues that platforms transition from a pure attention economy to a trust economy, actively protecting and highlighting human-made works to preserve the integrity and sustainability of the cultural value chain.