An answer engine is a search system that responds to a user query with a synthesized answer — usually generated by a large language model over retrieved web sources — rather than a ranked list of blue links.
Examples in 2026 include Perplexity, ChatGPT Search, Google AI Overviews, Google AI Mode, Bing Copilot, Gemini, and Claude with web search. The category also covers older systems that stitched short answers on top of retrieval — for example the answer boxes at the top of a classic Google SERP.
The mechanic is retrieval-augmented generation (RAG): the system searches the web, retrieves candidate passages, and passes them to the model with instructions to answer using only those passages and to cite each source. Perplexity's public FAQ describes exactly this loop, and Google's AI-features documentation confirms AI Overviews are generated on retrieved sources from the same web index that powers Search.
For publishers, the shift matters because success metrics change. A blue-link result earned a click; an answer engine may quote the source without the user visiting. That is the trade-off AEO is designed for: optimize for citation, not just for click.
Answer engines are not a replacement for classic search — they are a new surface on top of it. Both matter.