Against Opacity: Explainable AI and Large Language Models for Effective Digital Advertising

Digital advertising platforms lack transparency in how they target audiences and price ads, with algorithms using vast amounts of proprietary data unavailable to advertisers, resulting in billions of dollars lost on ineffective social media campaigns that rely on intuition rather than informed decision-making.

The paper proposes using machine learning and large language models (LLMs) to help marketers better understand advertising performance, specifically by predicting click-through rates (CTR) of ad content and summarizing competitive landscape data that would otherwise be too complex to analyze manually.

Standardized metrics and stronger regulatory frameworks are needed to address the opacity problem, as the current dominance of platforms like Meta and Google—reinforced by network effects—creates an imbalance where advertisers lack the transparency necessary to optimize their strategies effectively.

LLMs like ChatGPT can bridge the gap between complex advertising data and non-technical marketers, making it possible for advertisers to derive meaningful insights from large volumes of competitive ad data without requiring specialized technical expertise.

Explainable AI is positioned as a solution to responsible digital advertising, enabling both greater transparency in how ads are targeted and priced, and empowering advertisers to make data-informed decisions rather than relying on platform algorithms they cannot understand or verify.