Detecting Generated Native Ads in Conversational Search
This addresses the potential issue of subtle, manipulative advertising in conversational search for users, representing an incremental step in ad detection.
The paper tackles the problem of detecting automatically inserted ads in conversational search responses by compiling a dataset of queries and generated responses with ads and evaluating detection methods. The result shows that while LLMs struggle, sentence transformers achieve precision and recall above 0.9.
Conversational search engines such as YouChat and Microsoft Copilot use large language models (LLMs) to generate responses to queries. It is only a small step to also let the same technology insert ads within the generated responses - instead of separately placing ads next to a response. Inserted ads would be reminiscent of native advertising and product placement, both of which are very effective forms of subtle and manipulative advertising. Considering the high computational costs associated with LLMs, for which providers need to develop sustainable business models, users of conversational search engines may very well be confronted with generated native ads in the near future. In this paper, we thus take a first step to investigate whether LLMs can also be used as a countermeasure, i.e., to block generated native ads. We compile the Webis Generated Native Ads 2024 dataset of queries and generated responses with automatically inserted ads, and evaluate whether LLMs or fine-tuned sentence transformers can detect the ads. In our experiments, the investigated LLMs struggle with the task but sentence transformers achieve precision and recall values above 0.9.