Original title: Consumer responses to algorithmic in-feed advertising: AI transparency as a moderator among young adults in social commerce
Zhan Huang, Yan Li, Depeng Du · September 2026
Social media feeds use artificial intelligence to show you personalized ads, but researchers want to know how these ads shape our choices. To find out, researchers surveyed people about their reactions. They studied how ad features influence two internal reactions: affective attitude, which means how you feel emotionally about an ad, and cognitive attitude, which means what you logically think of it. The study found that fun, personalized ads with perks make people feel good, while useful, believable, and personalized ads make people think well of them. However, ads that feel intrusive damage both feelings and thoughts. When people have positive feelings or thoughts toward an ad, they are more likely to buy the item and share the post. Finally, knowing how the algorithm works strengthens the link between thoughtful opinions and buying things, but does not change how emotions drive purchases.
Imagine walking through a market where a vendor calls out to you. Some calls are funny (entertaining), some mention your name (personalized), some offer a discount (incentive). But if the vendor shouts the same facts at everyone (informativeness) or seems dishonest (credibility), you might not feel good even if the info is useful. If they block your path (intrusiveness), you get annoyed no matter what. In this paper each ad attribute is like a different sales pitch quality, and how you feel or think about the pitch determines whether you stop to buy or share the gossip with friends.
Young adults see hundreds of these AI-driven ads in apps like Instagram or TikTok every day. This research shows that not all ad qualities matter equally—entertainment targets feelings, while facts target thinking. For a platform to actually sell something, it has to know which button to push and how much transparency about AI to show.
The study only surveyed young adults in China, so results may not generalize to other age groups or cultures. It relies on self-reported intentions, not actual purchase behavior. The abstract does not report effect sizes or compare the relative strength of each path. AI transparency moderation on other relationships (e.g., affective attitude to purchase) was not found, leaving the full role of transparency unclear.
Ask anything about this paper. Answers come from its abstract, and say so when the abstract doesn't cover your question.
Consumer responses to algorithmic in-feed advertising: AI transparency as a moderator among young adults in social commerce — Zhan Huang, Yan Li, Depeng Du — Scientific Reports — doi:10.1038/s41598-026-72748-9
Abstract licence: CC BY
Explanation generated by Synthesis from the abstract — it is not the original paper.