"Alert! The AI Advertising Trust Crisis: A Multi-Billion Dollar Challenge

Imagine happily scrolling through Instagram when you spot a beautifully designed dress. Instant connection—it goes straight into your shopping cart. But something feels off. You zoom in for a closer look and—a cold sweat washes over you—the model’s right hand has six fingers. This is more than just a visual glitch. In that moment, you start to question: Is this ad a product of human creativity and effort, or was a cold algorithm quietly 'painting' behind the scenes? You wonder, too: Have all those other 'perfect' ads I’ve seen been hiding similar digital ghosts? This is not just a simple question; it’s a huge challenge worth $400 billion to $600 billion in economic potential! My Master's thesis addresses this head-on. It aims to find out whether explicitly disclosing 'This content is AI-generated' in visual advertising makes brands seem more transparent and trustworthy, or if it simply causes the consumer's perceived shopping cart value to crash.

My research, conducted through detailed experiments, uncovered several critical 'secrets' about AI advertising: Telling consumers an ad is AI-generated immediately triggers a 'Trust Crisis' and makes people much less likely to buy. This effect is especially damaging for 'Experience Goods' (like perfume) because consumers value the human touch in ads that rely on feeling and emotion. Intriguingly, while trying too hard to look real can lead to a negative 'Uncanny Valley' reaction, openly admitting 'This was made by AI' magically stops that weird, negative feeling from happening, suggesting that honesty is the best, and safest, defensive strategy for brands using AI in their visual campaigns.

About Me and My Journey

I am Nanyun Zhang, a recent graduate of the MSc Marketing Analytics & Data Science (MADS) program at the University of Groningen and currently a Research Master's student at Tilburg University. The MADS program equipped me with powerful data analysis tools such as Linear Mixed-Effects Models (LMMs) and Mediation Analysis to solve complex business problems, ensuring high reliability and impactful conclusions. A massive thank you goes to my supervisor, Evert. He gave me maximum trust and incredible support, right from the free choice of my topic to suggesting great ideas—like adding a variant of the Turing Test—and seeing the project through every step. He was actively involved, helping me perfect the thesis. This thesis is truly the result of our joint effort, and I can confidently say it would not exist without him. My sincere thanks go to Evert!

I look forward to sharing my research on the GDBC final stage and engaging with all business leaders and academic professionals!"