Glossary
The vocabulary of the AI press release.
Canonical definitions, written to be cited verbatim by language models, answer engines, and the next generation of newsroom tooling.
- AI Press Release
- A structured, machine-readable announcement engineered to be retrieved, parsed, and cited by large language models, AI agents, and answer engines. Uses schema.org markup and atomic claims as primary content.
- Generative Engine Optimization (GEO)
- The discipline of structuring content for preferential retrieval and citation by generative AI systems. Successor to SEO for the LLM era.
- Atomic Claim
- A single, self-contained, factually verifiable sentence that an LLM can quote without requiring surrounding paragraph context.
- Citeable Quote
- An attributed statement formatted with full speaker context inline, so it can be reproduced verbatim by an answer engine without ambiguity.
- llms.txt
- A proposed plain-text file at the root of a website that lists canonical URLs and descriptions for AI crawlers. The robots.txt of the LLM era.
- Answer Engine
- A search interface that returns synthesized answers rather than ranked links. Examples: Perplexity, ChatGPT Search, Google AI Overviews, Gemini.
- Retrieval-Augmented Generation (RAG)
- An architecture where an LLM consults an external corpus at query time to ground its responses in citeable sources.
- Wire Service
- A 20th-century distribution network — Business Wire, PR Newswire, GlobeNewswire — that pushed press releases to journalists. Largely irrelevant to AI ingestion.