Position Auctions in AI-Generated Content

2025-06-03

AI Summary

This paper addresses the challenge of naturally integrating sponsored content into AI-generated text, where contextual relevance and placement flexibility are critical. We propose the first auction mechanism for dynamic, context-sensitive textual placements—departing from static ad slots—by modeling click-through rate (CTR) at the fine-grained position-creative level. Our framework integrates LLM-driven contextual CTR estimation, multi-layer matching optimization, and dual user-behavior models: order-sensitive (Multinomial Logit, MNL) and order-insensitive (Cascade). Theoretically, we design an efficient and exact optimal mechanism for the MNL model; for the Cascade model, we provide a provably near-optimal solution with theoretical guarantees. Empirical results demonstrate significant improvements in the joint optimization of advertiser welfare and platform revenue, advancing the paradigm of AI-native advertising mechanism design.

Technology Category

• Multiagent Systems: Mechanism Design
• Search and Optimization: Learning to Search
• Game Theory and Economic Paradigms: Mechanism Design

Application Category

• Economics, Online Markets and Human Computation: Uses of LLMs and GenAI for marketplace design, bidding, and strategic interactions
• User Modeling, Personalization and Recommendation: User modeling for targeted and personalized online advertising
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

We consider an extension to the classic position auctions in which sponsored creatives can be added within AI generated content rather than shown in predefined slots. New challenges arise from the natural requirement that sponsored creatives should smoothly fit into the context. With the help of advanced LLM technologies, it becomes viable to accurately estimate the benefits of adding each individual sponsored creatives into each potential positions within the AI generated content by properly taking the context into account. Therefore, we assume one click-through rate estimation for each position-creative pair, rather than one uniform estimation for each sponsored creative across all positions in classic settings. As a result, the underlying optimization becomes a general matching problem, thus the substitution effects should be treated more carefully compared to standard position auction settings, where the slots are independent with each other. In this work, we formalize a concrete mathematical model of the extended position auction problem and study the welfare-maximization and revenue-maximization mechanism design problem. Formally, we consider two different user behavior models and solve the mechanism design problems therein respectively. For the Multinomial Logit (MNL) model, which is order-insensitive, we can efficiently implement the optimal mechanisms. For the cascade model, which is order-sensitive, we provide approximately optimal solutions.

Problem

• Extending position auctions to integrate sponsored creatives within AI-generated content contextually
• Estimating click-through rates for each position-creative pair using advanced LLM technologies
• Designing welfare and revenue-maximizing mechanisms for order-sensitive and order-insensitive user models

Innovation

• Integrates sponsored creatives into AI-generated content contextually
• Uses LLM to estimate position-creative pair click-through rates
• Formulates welfare and revenue-maximizing auction mechanisms