A June 2025 study revealed that Reddit was the most frequently cited web domain by large language models (LLMs), being referenced in approximately ** percent of analyzed instances. This prominence can be attributed to the content licensing agreement established between Google and Reddit in early 2024, which facilitated the training of AI models. Following Reddit, the second most cited domain was *** with mentions in roughly percent of cases. Additionally, the domains ** and ** were referenced in ** percent of the occurrences. This study underscores the significance of content partnerships and licensing agreements in shaping the training data for AI technologies. Incorporating reputable sources enhances the credibility and accuracy of LLM outputs, making this research pivotal for understanding the evolution of AI model training. Utilizing strong SEO practices, this summary highlights critical statistics and trends while optimizing for discoverability in search engines.
