AI Context Association (AICA): Exploring Context-on-Demand for the Agentic Web

AI Context Association (AICA) is an exploratory technical concept that investigates how individual web content elements could reference external, machine-readable context that AI agents can retrieve when needed. Modern websites are designed primarily for human users, often presenting only a subset of the information available in underlying content management systems, product platforms, and enterprise knowledge sources. Embedding all potentially relevant information directly into HTML may increase page payload, while conventional APIs and integrations can require additional implementation effort. AICA explores a lightweight association between a specific web element and a separate context resource, such as a webpage, document, or knowledge source. The objective is to investigate whether this approach can help AI agents access richer, relevant information without unnecessarily expanding the human-facing page. The concept is intended to complement existing technologies rather than replace established approaches such as Schema.org, JSON-LD, the Model Context Protocol (MCP), or WebMCP. Its potential value, interoperability, security implications, and performance benefits remain to be evaluated through implementation and comparative experimentation. AICA is presented as a research hypothesis and an invitation for technical review, not as an established web standard.