Calibrating Professional Discretion in Artificial Intelligence-Supported Business-to-Business Selling

Abstract

Abstract Artificial intelligence (AI) is becoming increasingly embedded in business-to-business (B2B) sales through applications such as lead scoring, forecasting, message generation, meeting summaries, customer relationship management, and next-best-action recommendations. Although these systems can improve speed and consistency, generative and emerging agentic AI also create new challenges for professional discretion because consultative selling depends on contextual interpretation, customer trust, ethical judgment, and the continuous development of salesperson expertise. This article presents a structured integrative review exploring how sales professionals and managers should calibrate their reliance on AI-generated sales recommendations. Drawing on research in AI-supported selling, human–AI reliance, automation bias, customer trust, adaptive selling, and sales management capability, the article develops a professional discretion calibration model. The model proposes four judgment filters (interpretive, relational, ethical, and developmental) that guide five response actions: accept, modify, reject, disclose, or escalate. The article contributes to sales management scholarship by explaining the judgment process through which AI-generated outputs are translated into customer-facing actions and by outlining the practical implications for the responsible use of AI in B2B sales.