By Nicole Penn Reddit accounted for an average 3.83 percent of ChatGPT Search citations from July 18 through August 7, according to AI visibility platform PromptWatch. By August 14, that share sat below 1 percent, averaging 0.52 percent through the following Monday. OpenAI has not explained the change, and PromptWatch calls its own finding “provisional.” Reddit in Google’s AI Overviews The drop is confined to one engine. Reddit continues to appear in Google’s AI Overviews and Gemini at normal rates, which points to something inside ChatGPT’s retrieval rather than anything about Reddit’s content. Lou Kelly at Boldspace tracked the same decline and described Reddit as cited less often rather than absent, with the pattern having begun around August 10. Still the Most-Cited Source Reddit also remains the most-cited source across AI engines in aggregate. The consolidated citation index published this year (built from more than 680 million citations across ChatGPT, Gemini, Perplexity, Claude, and AI Overviews) put Reddit at the top of the list at roughly 40 percent multi-engine frequency. A dip inside one engine does not move a brand out of that position. Comparisons Engine-level volatility is the normal condition of this channel. Reddit’s ChatGPT share swung by a similar magnitude last September and recovered. Wikipedia’s ChatGPT share moved by more than thirty points across the same period. LinkedIn climbed from around eleventh to fifth place between November and February. Any program that treats a monthly citation reading as a verdict will spend the next year chasing noise. So the recommendation for clients running Reddit programs has not changed. Community presence earns citations across Gemini, Perplexity, and AI Overviews regardless of what ChatGPT does this month, and it feeds a search-visible surface that has value with no LLM involved. Conclusion These changes can be hard to keep up with. EGC Group’s GEO audit maps citation share engine by engine rather than in aggregate, identifies the specific sources driving citations for your category, reviews how your owned content is structured for retrieval, audits entity accuracy across the encyclopedic and knowledge-panel sources the models treat as foundational, and shows where earned media is carrying visibility that community content cannot.
