Advertising efficiency is expected to improve as generative AI analyzes users’ situations and purchase intentions to deliver personalized ads. However, this is also fueling growing concerns about privacy and advertising transparency, with questions such as, “To what extent are the personal conversations I’ve shared being used for commercial purposes?”
As generative AI begins to function as a personal assistant beyond mere search, the competitive landscape in the advertising market is shifting from keyword matching to “analysis of conversational context and intent.”
Whereas traditional search ads connected search terms entered directly by users with keywords purchased by advertisers, AI ads infer the problems users are trying to solve and their potential needs from a series of conversations.
For example, if a user asks the AI how to buy groceries for a weekend house party, the AI can recommend relevant products by comprehensively evaluating not only the word “groceries” but also the number of guests, budget, and delivery schedule. Advertisers can present products tailored to the user’s situation from the earliest stages of the purchasing decision, and platforms can expand their advertising reach to stages even before the search takes place.
Projected to Grow 12-Fold Within 3 Years… The Race for AI Advertising Dominance Begins
Global Big Tech companies are exploring new revenue opportunities by integrating advertising models with generative AI services.
According to market research firm eMarketer, the U.S. AI search advertising market is projected to grow approximately 12.4-fold, from $2.08 billion in 2026 to $25.93 billion in 2029. During the same period, AI advertising’s share of the overall search advertising market is expected to expand from 1.3% to 13.6%.
Google is integrating search and shopping ads into its “AI Overview” feature, which uses AI to summarize search results, as well as into its conversational search services. Microsoft (MS) is also experimenting with displaying ads tailored to users’ search intent on the Copilot response screen.
Amazon is working to refine product recommendations by combining user queries with purchase data through its shopping AI, “Rufus.”
In South Korea, Naver is accelerating efforts to integrate generative AI search services with advertising. Naver is applying its AI advertising product, “AdBoost Max,” to “AI Briefing”—a feature that summarizes search results—to deliver ads tailored to search terms, the context of responses, and users’ browsing patterns.
It remains to be seen whether Naver, which has secured a competitive edge in the domestic advertising market based on search data, can maintain its existing advantage in the era of generative AI. However, some analysts suggest that if AI search replaces existing search traffic, changes to its search-advertising-centric business model will be inevitable.
OpenAI is also emerging as a wild card in the AI search advertising competition as it explores the potential for an advertising model based on ChatGPT. Lee Dong-jae, Account Director at OpenAI, explained at “Max Summit 2026” held in Seoul on the 3rd, “We are moving toward displaying ads by understanding the intent behind a conversation, rather than simply matching keywords.”
OpenAI’s policy is to refrain from displaying ads in sensitive conversations—such as those involving health, mental health, or emotionally vulnerable situations—and to exclude illegal or harmful content, as well as regulated products, from advertising. The company explained that an AI classifier determines whether an ad can be displayed, with human review conducted in parallel.
However, there is a limitation in that it is difficult for external parties to verify the criteria the AI uses to classify conversations and select ads.
“Analyzing Even Our Innermost Thoughts?”… Privacy Controversy in the Age of AI Advertising
What sets AI advertising apart from traditional search ads is its ability to infer information—even details users did not directly input—based on the context of the conversation.
While conversing with AI, users share information—such as family issues, financial difficulties, and job-related concerns—that they would not typically enter into a search bar. Even if the platform does not provide the original conversation text to advertisers, the fact that AI can analyze context to infer a user’s situation and interests could fuel growing concerns over privacy.
Another issue that sets this apart from traditional search ads is that it is difficult for users to verify why a specific ad was displayed or what information was used.
Plexicity AI tested ads in the form of sponsored questions displayed next to AI responses but halted the initiative due to concerns about undermining trust. Anthropic also announced its policy to operate its AI service “Claude” without ads, stating that the AI assistant’s judgment must not be influenced by advertisers’ interests.
Kim Myung-ju, director of the Artificial Intelligence Safety Research Institute, pointed out, “Generative AI can analyze even user tendencies that the users themselves are unaware of,” adding, “We need principles that can explain what information is used for advertising and why a specific ad was displayed.”
The key to success in the AI advertising market is expected to go beyond the sophistication of recommendations to include transparency in data usage and securing user trust. Analysts suggest that platforms where users feel comfortable entrusting their conversations—rather than those with the largest amounts of data—are more likely to gain a competitive edge.
