Det känns... AI-genererat : En kvalitativ studie om hur upplevt AI-innehåll påverkar användares förtroende för företag

Sammanfattning

Technically flawless, factually accurate, yet somehow untrustworthy. This is the paradox users face when perceiving corporate content as AI-generated. This study investigates how AI-generated content in corporate digital communication affects trust and attitudes toward companies. The study applied a qualitative approach with an inductive stance, combining survey data and semi-structured interviews. Four themes emerged. First, human presence is key to perceived authenticity, so much so, that imperfections can function as credibility markers, signalling that a real person was involved – someone who cared, or at the very least reviewed the content. Second, users find transparency about AI use necessary, but insufficient on its own. The transparency must be combined with proof of human involvement and accountability. Third, what counts as “acceptable AI use” is not universal, but deeply contextual. The findings point to two dimensions shaping these judgements: the resources of the company, and the moral weight of what is being communicated. The fourth and most striking theme: a single perceived AI signal can be enough to trigger what this study terms a spillover effect: users do not evaluate content in isolation, but instead reassess the company as a whole: its values, its priorities, and its relationship with them. These findings suggest that generative AI use in digital communication is fundamentally a UX concern with trust- and relational consequences for companies, not a question of efficiency alone.