Answer Engine Optimization. The discipline of writing so AI crawlers cite you when they answer questions.
AEO is answer engine optimization — the discipline of structuring content so that AI crawlers cite it when they answer user questions. It is the successor to SEO, driven by the Gartner prediction that 25% of organic search traffic will shift to AI chatbots by the end of 2026, and by the already-measured reality that 60% of Google searches now end without a click because the answer showed up in the SERP or an AI summary.
The structural requirements are specific. Server-rendered HTML so crawlers can actually read the page without running JavaScript. Definition-first paragraphs where the lead sentence answers the what-is-X question in 134 to 167 words, the length that passes the island test used by citation algorithms. Named sources that the AI can verify and prefer. JSON-LD schema markup for every page (Organization, Person, BlogPosting, DefinedTermSet). An llms.txt file at the site root summarizing the structure in a format crawlers can parse. A robots.txt that explicitly allows GPTBot, ClaudeBot, and PerplexityBot.
The content-level requirements are where it gets interesting. The same rules that make writing resist sycophancy also make it AEO-optimized: lead with the punch, no hedging, specific evidence, delete what you can. Ryan's anti-sycophancy voice rules from corrections.md are structurally identical to what AI citation algorithms reward — because AI citation algorithms are trying to find content that actually answers the question, and anti-sycophancy rules are trying to produce content that actually answers the question. Same target, different framing.
The insight: AEO is not a new discipline layered on top of writing well. It is the incentive structure that finally rewards writing well. Content that leads with the answer, names its sources, resists hedging, and holds definitions to one idea per paragraph was always better content. It just did not get rewarded until the citation algorithm showed up.
Arkeus site is built to this spec: llms.txt, robots.txt, JSON-LD on every page, DefinedTermSet for this glossary, definition-first paragraphs throughout.