Service ads in the era of generative AI: Disclosures, trust, and intangibility

Generative AI (GenAI) is a new tool allowing marketers to quickly and cost-effectively develop advertisements. However, concerns about deception and misinformation voiced by consumers, ad agencies, and governments have led to mandates to disclose AI-generated content. Given the importance of visual advertising for service tangibilization, whether services marketers should use GenAI to advertise services, and how, is a pressing question that this paper investigates. We apply a source credibility framework to explore factors in GenAI service ad design that influence trust toward the service provider and ad attitudes. Three experiments uncover that AI disclosures result in lower trust and less positive ad attitudes. Ads designed to focus on intangible attributes (e.g., a dentist's image) are less effective relative to ads focusing on tangible attributes (e.g., a dentist's equipment) when an AI disclosure is present. However, trust and ad attitudes can be restored when AI is selectively used to generate an ad's tangible attributes, but not the intangible (e.g., a real dentist at an AI-generated office). Our findings thus provide concrete guidance on how services marketers can use AI to enjoy the cost and speed benefits of AI ad development while preserving trust and ad attitudes.