Agentic AI Optimisation (AAIO): what it is, how it works, why it matters, and how to deal with it

AAIO is a nascent conceptual framework introduced via an ArXiv paper just days ago. While it correctly identifies a massive shift—from Human-to-Machine (H2M) optimization (SEO) to Machine-to-Machine (M2M) optimization—it currently lacks any technical moat or proprietary implementation. The defensibility is low (2) because it is a thought-leadership piece rather than a functional tool or protocol. The 7 forks within 3 days indicate high interest in the 'SEO for Agents' category, but there is no code to prevent replication. Frontier risk is high because the labs building the agents (OpenAI, Google, Anthropic) will likely dictate the standards for agent-web interaction (e.g., extensions of robots.txt or specific JSON-LD schemas). If Google Search Generative Experience (SGE) or OpenAI's Operator defines a standard, AAIO must align or face irrelevance. The primary threat is platform domination: web standards for agents will likely be consolidated by the browser and LLM providers rather than third-party frameworks. Comparable to early 'LLM Optimization' papers, this project's value lies in its taxonomy rather than its technical execution at this stage.