University of Hong Kong Researchers Xing Hu and Jitong Yao Release New Research on AI Search Answers as Market Gateways

A large-scale University of Hong Kong audit finds generative AI answers may shape which service firms consumers see first.

HONGKONG, HONG KONG, July 6, 2026 /EINPresswire.com/ -- University of Hong Kong researchers Xing Hu and Jitong Yao have released new research examining how generative AI search may be changing the way consumers discover service providers.

The study finds that AI-generated answers are becoming a new market gateway, influencing which companies are seen, compared and considered at the earliest stage of consumer decision-making. Unlike traditional search engines, which display multiple links and ranked results, generative AI systems often compress a market into a short synthesized answer. That shift means firms are no longer competing only for search rankings or clicks. They are also competing to be included in the answer itself.

The research analyzes 36,000 AI-generated answers across 10 service industries, four model families, 200 query scenarios and 45 observation rounds. The research team built 260,372 deduplicated answer-firm visibility records and identified 7,250 verified firm entities appearing in AI-generated service recommendations.

“Generative search does not simply retrieve information,” the researchers write. “It constructs a representation of the market.”

The findings show substantial concentration in answer-layer visibility. Across the audited firms, cumulative visibility had a Gini coefficient of 0.888. The top 1% of firms captured 45.1% of total visibility, while the top 5% captured 73.7% and the top 10% captured 84.0%.

The study also found that early visibility advantages tended to persist. Among firms that appeared in the top visibility group during the early observation period, 85.6% remained in that group later. By contrast, only 1.61% of firms that were not initially in the top group later moved into it.

The study pays particular attention to small and medium-sized enterprises. Large firms recorded an average cumulative visibility score of 17.57, compared with 2.09 for small and medium-sized firms. Large firms were also more likely to appear in top answer positions, with a top-position rate of 27.6%, compared with 6.14% for small and medium-sized firms.

The researchers emphasize that the findings should not be read as evidence that AI systems intentionally exclude smaller businesses. Instead, the study points to a structural issue: generative AI systems may be more likely to identify, verify and summarize firms with clear public information, consistent naming, structured profiles, third-party references and stronger digital footprints.

A central concept in the study is “answerability,” which refers to the extent to which a firm can be recognized, cross-checked and converted into a credible answer by an AI system.

“For small businesses, the challenge is not only digital marketing,” said Jitong Yao. “It is becoming answerable to AI systems. A company needs public information that can be recognized, verified and explained.”

The study has implications for consumers, businesses and policymakers. For consumers, AI-generated recommendations may appear comprehensive while representing only a compressed selection of available providers. For businesses, especially smaller firms, online discoverability may increasingly depend on the quality, consistency and verifiability of their public information footprint.

The researchers suggest that future AI search governance should look beyond misinformation, bias and ranking fairness. It should also examine which firms are included in AI-generated answers, which firms are omitted, how recommendations are justified and whether smaller providers have realistic pathways to become visible.

The research contributes to ongoing discussions about generative AI, platform governance, small business digitization and consumer decision-making. It suggests that the future of online competition may depend not only on being searchable, but also on being answerable.

jitong yao
University of Hong Kong
email us here
Visit us on social media:
LinkedIn

Legal Disclaimer:

EIN Presswire provides this news content "as is" without warranty of any kind. We do not accept any responsibility or liability for the accuracy, content, images, videos, licenses, completeness, legality, or reliability of the information contained in this article. If you have any complaints or copyright issues related to this article, kindly contact the author above.