Computational Models for Public Trust in Emerging Technologies

Abstract

Public trust in emerging technologies -- AI systems, autonomous vehicles, biotechnology, and digital infrastructure -- is both a prerequisite for beneficial technology adoption and a product of governance quality, past technology experiences, and media framing. Computational models of public trust dynamics enable policymakers and technology developers to understand how trust forms, erodes, and recovers; to predict the adoption implications of trust crises; and to design governance interventions that build warranted trust. This paper proposes the Computational Public Trust Model for Emerging Technologies (CPTMET), applying system dynamics and agent-based modelling to simulate public trust dynamics across four technology domains: AI personal assistants, autonomous vehicles, CRISPR gene therapy, and smart grid infrastructure. CPTMET introduces the Trust Dynamics Quality Score (TDQS) assessing model validity, policy sensitivity, and predictive accuracy. Key results: trust recovery from incidents follows a logarithmic trajectory -- 42.4% of pre-incident trust recovers within 24 months with no governance response, rising to 78.4% with active governance intervention; media framing has a 2.8x amplification effect on incident trust impact; transparent governance is the single most effective trust-building intervention (+28.4% trust gain per unit investment). The framework provides trust-aware technology governance guidance for policymakers and deployers.