An essay on trust as economic infrastructure argues that deepfakes and AI-driven automation are undermining confidence in identities and outputs across sectors, creating new friction for compliance and transactions. It cites examples including a deepfake-driven fraud at Arup that led to a $25.6 million loss after synthetic executive video calls, and other failures where AI-assisted outputs included fabricated sources and incorrect inventory. The piece frames trust as something firms must engineer—covering facts, authentication, system security, and accountability when errors occur. It links the breakdown to slower decision-making, higher insurance and compliance costs, and reduced risk-taking. For higher education, the implications cut across research integrity, procurement and vendor verification, and student and staff data systems increasingly exposed to synthetic content threats.