The proliferation of large language models (LLMs) is fundamentally transforming the production, distribution and consumption cycle in the digital media ecosystem. The motivation for this research was to examine this transformation not only from a technology perspective, but also from the standpoints of media economics and user behavior. Through a systematic literature review and case analyses covering the 2023-2024 period, we evaluated the role of models like GPT-4, Claude and Gemini in content production processes and their impact on journalistic practices.
Methodologically, we adopted a three-layer framework: the production layer (content creation, summarization, localization), the distribution layer (personalization algorithms, SEO automation), and the reception layer (reader trust, disinformation risk, media literacy). For each layer we compiled concrete case studies from the literature on how LLMs are changing existing practices; we examined the AI integration experiences of media organizations such as AP, Reuters and BuzzFeed.
Findings revealed that LLMs provide high efficiency particularly in routine news content with low added value (financial reports, sports statistics, weather). In contrast, investigative journalism, context-requiring analytical content and local news production were found to remain critically dependent on human editorial judgment. Regarding disinformation risk, we emphasized that content verification mechanisms need to evolve alongside LLM integration, and that model watermarking and provenance tracking systems will be a core research area in the coming period.
The most fundamental conclusion from this research is that the question 'will AI replace humans?' has already been surpassed. The real question is at which layer and with what responsibility distribution the human-AI collaboration should be structured. Since the research was written with an interdisciplinary perspective, it aimed to contribute to both technical and social science literature. Access figures on ResearchGate confirmed that the intersection of AI and media is attracting increasing interest among academics.