This study investigates whether ChatGPT-driven traffic (CGT) acts as a substitute or complement to traditional news media platforms in the United States and Taiwan. By analyzing website traffic patterns and incorporating cross-cultural and platform-specific factors, the research aims to determine how generative AI tools influence news consumption dynamics. The study also examines the moderating effects of website scale and specialization (generalist vs. niche) to provide a nuanced understanding of AI’s role in reshaping digital media ecosystems. Findings offer insights for media organizations, policymakers, and AI developers on adapting to evolving content distribution and engagement trends.
This study employs a quantitative research approach using Partial Least Squares Structural Equation Modeling (PLS-SEM) to examine the impact of ChatGPT-driven traffic (CGT) on total website traffic (TWT). A dataset of 80 news websites (40 from the U.S. and 40 from Taiwan) was analyzed, incorporating website scale and specialization (generalist vs. niche) as moderating variables. Traffic data were sourced from SimilarWeb, capturing monthly visits over six months. Multi-group analysis (MGA) was conducted to compare effects across different platform sizes and national contexts, offering insights into ChatGPT’s role as a substitute or complement in news media ecosystems.
The study reveals that ChatGPT-driven traffic (CGT) generally complements news media by increasing website visits, but its impact varies by country and website scale. In Taiwan, ChatGPT acts as a traffic driver, especially for smaller and niche platforms. In contrast, large U.S. websites experience substitution effects, suggesting AI-driven content may reduce direct visits. Website scale is a key factor, with smaller platforms benefiting more from AI referrals. These findings highlight the need for adaptive strategies in media organizations and call for regulatory considerations to balance AI’s role in content distribution and news ecosystem sustainability.
This study relies on third-party traffic data (SimilarWeb), which may not fully capture user interactions and engagement depth. The analysis is limited to a six-month period, potentially missing long-term trends in AI-driven news consumption. Additionally, the findings focus on the U.S. and Taiwan, limiting generalizability to other media markets with different regulatory and technological landscapes. The study examines traffic patterns but does not explore user trust, perception, or content credibility in AI-generated summaries. Future research should incorporate longitudinal data, qualitative insights, and additional geographic contexts to provide a more comprehensive understanding of AI’s impact on news ecosystems.
News organizations, especially smaller and niche platforms, should optimize content for AI-driven referrals to enhance visibility and engagement. Large U.S. platforms facing substitution effects must focus on exclusive content and direct user engagement to maintain traffic. Policymakers should ensure transparent referral attribution and explore fair revenue-sharing models between AI platforms and news publishers. AI developers should collaborate with media outlets to create sustainable content distribution frameworks, balancing efficiency with news ecosystem sustainability. Understanding ChatGPT’s role in media consumption can help stakeholders adapt strategies, refine monetization models, and foster AI-media partnerships for long-term industry resilience.
The rise of AI-driven news consumption impacts public access to information, media trust, and digital literacy. In Taiwan, ChatGPT enhances access to niche content, potentially fostering information diversity. In the U.S., substitution effects may contribute to reduced engagement with primary news sources, affecting media sustainability and journalism quality. AI-generated summaries could influence public opinion formation, raising concerns about bias and misinformation. Policymakers must promote transparency and ethical AI use to ensure balanced news dissemination. Media literacy initiatives are essential to help users critically engage with AI-curated content while preserving credible journalism and democratic discourse.
This study provides a novel cross-cultural analysis of ChatGPT-driven traffic (CGT) in the U.S. and Taiwan, offering new insights into AI’s role in news consumption dynamics. By integrating website scale and specialization (generalist vs. niche), it refines the substitution-complementarity framework in the context of generative AI. Unlike prior research focusing on social media and search engines, this study uniquely examines AI-driven referrals using empirical traffic data. The findings contribute to digital media strategy, policy discussions, and AI adoption research, guiding news organizations, regulators, and AI developers in navigating the evolving AI-media ecosystem.
