Reputation Arbitrage and the Crisis of Calibrated Trust: Defending the Public Against AI-Enabled Synthetic Deception

Abstract

The proliferation of generative artificial intelligence has precipitated a fundamental restructuring of the epistemology of trust in digital communication. This paper introduces the concept of “reputation arbitrage”, the systematic exploitation of borrowed credibility signals through synthetic media, cloned voices, and algorithmically generated impersonation, to describe an emerging class of socio-technical attacks that sever recognition from authenticity. Drawing on forensic linguistics, cognitive security theory, and behavioural economics, the paper argues that conventional cybersecurity paradigms, which privilege technical detection and credential verification, are insufficient when adversaries manipulate the decision layer itself: the human cognitive process by which trust becomes action. The paper extends the construct of calibrated trust, established in the human-automation trust literature (Muir, 1987; Lee & See, 2004), from the domain of machine reliability to the domain of interpersonal identity verification, and integrates this extension with an empirically grounded account of how time pressure and social isolation degrade deliberative judgement. Through an integrative analysis of regulatory frameworks, independently corroborated threat intelligence, and cognitive-architecture research, the paper proposes a five-step behavioural protocol (TRUST) designed to operationalise calibrated trust as a public epistemic defence, and situates this protocol within the existing landscape of scam-prevention mnemonics rather than presenting it as a discovery without precedent. The analysis establishes that the public is not the “weakest link” in cybersecurity but its constitutive decision layer, and concludes by outlining the theoretical foundations for the institutional and technological countermeasures developed in subsequent papers in this series.