Addressing vulnerability in customer experience with AI-agents

Purpose Artificial intelligence (AI) agents have the potential to fundamentally change customer experience (CX) by addressing vulnerability. Advancements in AI offer service accessibility and responsiveness, providing scope for improved customer experience. Any consumer can be vulnerable, depending on context, and as AI-agents proliferate, there is a risk that these agents prioritize interests of more powerful service providers over consumers. Instead, we envision personal AI-agents advocating on behalf of consumers, improving access to services, enhancing CX and shifting power from providers to consumers. Design/methodology/approach This conceptual theory-adaptation paper first synthesizes literature on consumer vulnerability to advance understanding of vulnerability in customer experience (CXV). Agency theory is then applied as a theoretical lens, building on the notion of perceived control applied to design personal AI-agents to act on behalf of consumers experiencing vulnerability. Findings Drawing on vulnerability, CX and AI in service literature, this paper, first, proposes a novel CXV conceptual framework to design personal AI-agents to enhance CX. Our conceptual framework identifies five design attributes that personal AI-agents should exhibit, underpinned by rebalancing perceived control from service providers, handing agency back to consumers. Second, we provide an actionable design framework comprising four archetypes to guide practice. Third, a compelling research agenda is offered to guide future research on addressing vulnerability in customer experience with personal AI-agents. Practical implications This paper provides practical guidance for agnostic third-party designers to develop personal AI-agents that can rebalance service provider interests with consumer advocacy. To design personal AI-agents, we propose four design archetypes in our 2x2 design framework: (1) service orchestrator, (2) protective sentinel, (3) reliable intermediary and (4) autonomous ally based on the five design attributes, tailored to meet the needs and preferences of consumers experiencing vulnerability. Originality/value This paper summarizes consumer vulnerability literature, developing the definition to provide a foundation for our five design attributes of personal AI-agents depicted in our C×V conceptual framework. This is integrated with practical illustrative vignettes of experiences, bodies, processes and regulations relating to real-world value delivery by AI-agents on behalf of consumers in service settings.