The first independent measurement of how AI engines cite the wire services. GlobeNewswire 31.3%. Business Wire 18.8%. PR Newswire 16.7%. Newsfile and ACCESS Newswire 12.5% each — but both win their target vertical outright.
Six academic studies that define how ChatGPT, Claude, Perplexity, Gemini, and Google AI Overviews cite the web. Authors named, methods explained, every number sourced. The reference document for practitioners, researchers, and journalists in AI visibility.
AI engines have replaced demographic audience segmentation. ChatGPT, Claude, Gemini, and Perplexity group buyers by the prompts they type — surfacing 14 primary consumer segments that are intent-shaped, not demographic.
The discovery layer has collapsed into five answer engines — ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews. Side-by-side data, citation behavior, and the linked playbook for each.
The University of Toronto's definitive audit of GPT-4o, Claude, Perplexity, and Gemini versus Google Search across 1,516 queries. Domain overlap, source typology, content freshness, and the controlled experiment on pre-training versus retrieval bias.
When an AI engine answers a buying question, the brand it names just ran an ad. The brand it omitted paid for the impression. Welcome to AEO.
Citation Share shows who gets mentioned. Agent Share shows who gets picked. In agentic AI buying, that is the only share that matters.
From Google operators (site:, inurl:, intitle:) to AI engine prompt mining. The 2026 SEO and GEO discovery research stack.
RAG — the architecture behind ChatGPT browsing, Claude web search, Perplexity, Google AI Overviews. Why published content is now retrieval inventory, why GEO is the discipline, and what comms teams must operationalize.