brand.context: A Machine-Readable Brand Context Standard for Agentic Commerce

Abstract This paper defines brand.context — a machine-readable brand context standard designed for consumption by AI agents during commerce and purchase recommendation tasks. brand.context is the implementation layer of the AIVO Evidentia Decision-Stage Filter Taxonomy (WP-2026-01), translating eight evidence-grade filter types derived from 7,000+ four-turn AI buying sequences into a structured JSON-LD schema that brands publish at a predictable location on their own domains. We identify a structural gap in the current AI commerce infrastructure: while AI agents are increasingly deployed to execute purchase decisions on behalf of consumers, no standard exists for brands to declare their decision-stage positioning in a format agents can reliably consume. Current brand content architectures were designed for human readers and search engine crawlers — not for agents that need to evaluate structured claims against specific criteria at inference time. brand.context addresses this gap by providing a schema in which every field maps directly to a filter type in the Evidentia taxonomy. A brand that publishes a correctly structured brand.context file is declaring, in machine-readable form, its position against the criteria that AI models have been empirically shown to use when eliminating brands at the decision stage of buying conversations. The standard is open, citable, and grounded in the only published evidence-grade taxonomy of AI decision filters derived from live multi-turn transcript analysis.