AdCP – the open standard for Agentic Adverising (github.com)

AdCP (Ad Context Protocol) is a newly announced open standard for advertising automation built on the Model Context Protocol (MCP). It provides a single, unified API that lets AI assistants (e.g., GPT, Claude) manage campaigns across DSPs and data platforms via natural language: discover audience “signals” (e.g., “high-income millennials interested in sustainable fashion”), compare pricing, activate segments, track real-time deployment, and receive standardized analytics. The spec and reference code (including a full MCP server example and Scope3 Signals Activation support) are available under Apache 2.0 at adcontextprotocol.org, with a validation test suite, docs for signals/media-buy/curation, and community governance via monthly working groups and GitHub discussions. For AI/ML and ad-tech teams this matters because AdCP standardizes signal schemas, reporting, and activation flows, reducing integration overhead and vendor lock-in while enabling programmatic, assistant-driven workflows. Technically, it formalizes multi-dimensional signal definitions (geo/temporal/behavioral), transparent CPM/revenue-share models, cross-platform deployment semantics, and real-time status APIs—making it easier to implement once and operate across platforms. Early adopters can implement an MCP-compatible server, run the test suite, and configure assistants to call AdCP endpoints; additional platform implementations are expected in Q1 2025, accelerating interoperable, AI-powered media buying and analytics.