• Image research suggests majority of people unable to distinguish between real and AI generated images

• AI images could have negative impact on trust – but less so with younger consumers and men

• AI images considered more likely to “pique curiosity”, but equally "captivating” and likely to “influence purchase” as real images

• Authentic, user generated content shown as key to building customer trust and purchase confidence

New research by TUI Musement, the Tours & Activities division of TUI Group, suggests the majority of people are unable to distinguish between real images and those generated utilising artificial intelligence (AI). The independent study, conducted by Appinio on behalf of TUI Musement in June 2024 with 1,200 adults across all age groups in Germany, Spain and the UK, also found that AI images could have a negative impact on peoples’ trust, although younger consumers and men seem more positive on the topic.

The research was carried out as TUI ensures the ethical implementation of AI across the business, and that any application enhances the customer or colleague experience.

“As AI evolves, it presents opportunities for the travel industry to enhance customer engagement. However, the importance of ethical AI deployment is paramount, and this type of research is key to understanding how best to integrate advanced technology while maintaining customer confidence,” said Peter Ulwahn, CEO of TUI Musement. “The findings from this research underscore the importance of balancing innovation with transparency. We must ensure that AI enriches the travel experience without compromising trust and our approach will remain rooted in authenticity."

Key Insights from the study:

• AI vs Real Images: 22% of participants from the UK said they can “easily distinguish” between AI and real images, but this falls to 9% in Germany and Spain. On average Gen Z and Millennials were more likely to claim they can “easily distinguish” between the two (17%), compared to Gen X (11%), and Baby Boomers (6%). However, actual tests revealed that all groups struggled to accurately differentiate between AI-generated images and real images.

• Trust Impact: While AI-generated images prompted feelings of "interest" in many participants, they also led to widespread “scepticism”. Spanish respondents, Baby Boomers and women express more cautioun when it comes to AI-generated imagery, whereas those from the UK, as well as Gen Z, younger Millennials and men, were more positive.

• Influence on Purchasing Decisions: Although AI-generated images were twice as likely to "pique curiosity," they were not considered more captivating or persuasive in driving purchase decisions. Spanish consumers were especially wary of AI’s influence, with most believing AI-generated images would negatively impact their purchasing decisions. In contrast, respondents from the UK and Germany expressed either neutral or positive attitudes.

• Preferred Imagery: When asked about online purchases, the majority of participants stated a preference for a blend of “high-quality professional photos” and “authentic, user-generated content”. This mix, especially with user-generated imagery, was identified by 88% of respondents as crucial in building trust and influencing purchasing behaviour.

TUI does not use AI-generated destination images on its platforms today. In cases such as TUI’s AI-generated influencer Lena, the destinations shown in the background are always real images. Furthermore, whenever AI is used, TUI clarifies this.

The research contributes to a better understanding for potential future use cases. TUI already uses AI technologies to automate and improve the management and optimisation of image content, classifying and filtering the more than 35 million images in its database, primarily hotel and room images. AI identifies and evaluates images based on technical and aesthetic criteria in order to make the best possible image selection for automated galleries and avoid duplicate or inappropriate images. It also supports keywording and categorisation by automatically assigning tags such as “summer”, “winter” or “pool” to images.

Future AI applications at TUI could include personalised descriptions, target group-relevant content and interactive travel advice for customers.