The AgenticAdvertising.org trade association provides a governance structure where brands, agencies, and publishers hold equal board weight regardless of company size. As the digital marketing landscape shifts from manual campaign management toward autonomous agentic workflows, the industry faces a significant hurdle: the lack of a common language for machines to negotiate and execute ad buys. Without a standardized framework, an artificial intelligence assistant tasked with launching a product campaign would require thousands of bespoke connectors to interact with the myriad of publishers, supply-side platforms, and data providers that comprise the modern ecosystem. The Ad Context Protocol, often abbreviated as AdCP, has emerged as a critical open technical standard to resolve this fragmentation by defining how software agents built on large language models buy and sell advertising. Rather than moving money or serving ads directly, the protocol acts as a sophisticated vocabulary of named operations, described in JSON schemas, that agents call over existing transport layers like the Model Context Protocol or Google’s Agent2Agent protocol. This shift ensures that a buyer’s agent can communicate seamlessly with any seller’s agent, regardless of the underlying ad server or proprietary technology stack being utilized, effectively laying the groundwork for a truly automated future in media procurement.
1. Defining the Core Protocol Components
The technical specification of the Ad Context Protocol is meticulously organized into several protocol surfaces, each designed to handle a specific facet of the advertising lifecycle. At the center of these operations is the Media Buy surface, which facilitates inventory discovery, campaign creation, and granular delivery reporting through commands such as get_products and create_media_buy. Complementing this is the Creative surface, which provides the necessary tools for building, previewing, and synchronizing advertisements across various platforms. By standardizing these interactions, the protocol allows creative agents to focus on asset generation while ensuring that the technical handoff to the delivery systems remains flawless. Additionally, the Signals surface manages the activation of targeting data and audience segments, allowing for sophisticated audience matching without the need for manual data uploads or fragmented API calls. These surfaces collectively ensure that every stage of an ad campaign, from initial planning to final creative synchronization, is handled through a unified interface that prioritizes interoperability over proprietary lock-in.
Beyond the immediate transactional tasks, the protocol also incorporates layers for identity, billing, and real-time execution. The Accounts surface oversees commercial identities and billing usage reporting, ensuring that financial tracking is as automated as the media buying itself. This is supported by the Governance and Intelligence surface, which introduces suitability standards and the brand.json advertiser identity file, a critical component for maintaining brand safety and transparency in an agent-driven world. For high-speed operations, the Trusted Match surface serves as a dedicated layer designed for impression-time activation across diverse environments like connected TV, mobile apps, and retail media. While traditional planning tasks might operate over several seconds or minutes, Trusted Match aims for a service-level target of roughly 50 milliseconds to keep pace with the demands of modern digital auctions. By integrating these diverse capabilities into a single framework, AdCP provides a comprehensive toolkit that addresses both the long-term strategic needs of marketers and the immediate technical requirements of real-time ad serving.
2. Execution of a Media Buy: A Step-By-Step Guide
Executing a transaction within the Advertising Context Protocol follows a logical progression that begins with the submission of a campaign brief. The buyer agent initiates the process by sending a get_products request to the seller agent, which includes a natural-language description of the goals and a specific buying_mode field. In the current version of the protocol, this field offers three distinct paths: the brief mode, where the publisher curates recommendations based on the input; the wholesale mode, where the buyer requests a raw product feed for independent audience application; and the refine mode, which allows for iterative adjustments to previous responses. Once the brief is received, the seller agent reviews the requirements and responds with a list of available inventory products. These recommendations include a range of pricing models, such as cost-per-thousand impressions, cost-per-click, or even revenue share and flat rates for fixed-price or auction-based environments. This standardized discovery process eliminates the friction of manual negotiation, allowing agents to evaluate thousands of potential opportunities in a fraction of the time required by traditional methods.
Once a suitable product is identified, the buyer agent proceeds to initiate the media purchase by calling the create_media_buy operation. This request is highly structured, requiring an account identifier, brand details, specific start and end times, and a mandatory idempotency key to prevent accidental duplicate bookings during network retries. Each package within the purchase names a specific product and pricing option, often accompanied by parameters for budget, pacing, bid price, and frequency caps. Following the booking, the buyer agent utilizes the sync_creatives task to transfer and link advertising assets to the campaign, ensuring that the creative content is ready for delivery. Finally, as the campaign runs, the agent continuously monitors performance by requesting data through the get_media_buy_delivery command. This provides real-time feedback on impressions, clicks, and other key performance indicators, allowing the buyer agent to report back to the advertiser or make further optimizations. This systematic approach ensures that every phase of the buying cycle is documented, verifiable, and executable across any platform that adheres to the protocol standards.
3. Key Participants and Governance Framework
The operational ecosystem of the protocol relies on clearly defined roles for participants on both the supply and demand sides of the market. On the sell side, publishers and supply-side platforms are responsible for running sales agents that interface directly with their inventory and ad-serving systems. To maintain a secure environment, publishers utilize a specialized adagents.json file to publicly declare which agents are authorized to represent and sell their digital space. This mechanism is increasingly viewed as a necessary successor to the older ads.txt standard, providing a more robust way to manage agentic permissions in a decentralized landscape. The technical maintenance of the reference sales agent is currently under the stewardship of Prebid.org, which ensures that the open-source code remains accessible and updated for the broader publishing community. By offloading the complexity of agent management to a trusted industry body, publishers can focus on optimizing their yield and audience insights while remaining compatible with the growing wave of automated demand from agentic buyers.
On the buy side of the equation, advertising agencies, demand-side platforms, and independent technology vendors operate buyer agents designed to seek out and acquire media on behalf of brands. These agents are programmed to parse natural-language briefs and translate them into the technical parameters required by the Ad Context Protocol, acting as the primary bridge between human strategy and machine execution. Oversight for the entire protocol is provided by AgenticAdvertising.org, a nonprofit trade association incorporated in Delaware. This body is structured to ensure that all stakeholders have an equal voice, preventing any single technology provider from exerting undue influence over the standard’s development. The governance model specifically targets the high integration costs that have historically plagued the advertising industry, seeking to replace hundreds of bespoke connections with a single, shared interface. This collaborative approach is essential for scaling the protocol beyond its founding members, as it creates a neutral ground where competitors can cooperate on the underlying infrastructure of the next generation of digital media trading.
4. Historical Timeline and Evolution of the Standard
The development of the Ad Context Protocol began in July 2025 when Brian O’Kelley made the initial code contribution to the project’s public repository. By October of that year, the protocol officially launched as version 2.0.0, backed by a founding group that included notable names such as Yahoo, PubMatic, and Triton Digital. These early adopters recognized the urgent need for a standardized communication layer as AI-driven assistants began to play a larger role in media planning and execution. In December 2025, the industry witnessed the first successful agent-to-agent transaction test, which involved Scope3 acting as the buyer agent and Magnite’s SpringServe platform serving as the seller agent. This milestone demonstrated that the theoretical framework could indeed handle real-world advertising tasks, from discovery to delivery. Since those early experiments, the protocol has moved from a niche experiment to a more mature standard, gaining traction among a broader group of participants who are eager to reduce the engineering overhead associated with managing diverse advertising APIs and proprietary platforms.
As the protocol evolved, the release of version 3.0.0 in April 2026 introduced significant structural changes and security enhancements designed to meet the demands of a changing regulatory environment. This version made idempotency keys mandatory for every mutating request, drastically reducing the risk of duplicate transactions and improving the reliability of high-volume ad buying. It also introduced specialized schema rules that enforced human review for sensitive advertising categories, such as housing, lending, and pharmaceuticals, in direct response to the EU AI Act and GDPR requirements. The older 2.x versions were fully deprecated by August 2026, signaling a total transition to the more robust and secure framework. Recent updates, including the version 3.3.0 beta released in October, have continued this trend by adding experimental agent identity schemes based on Web Bot Auth. These advancements reflect a commitment to continuous improvement, ensuring that the protocol remains resilient against fraud while providing the transparency that modern advertisers demand. By strictly managing the lifecycle of each task, the developers have created a predictable and scalable environment for autonomous agents to operate within.
5. Industry Impact and Strategic Future Considerations
Despite its technical strengths, the Ad Context Protocol faces significant challenges related to industry adoption and competition from alternative standards. A primary hurdle is the absence of major platforms like Google, Amazon, and Microsoft from the founding roster, as these companies often prefer to maintain their own proprietary APIs and walled gardens. Furthermore, the IAB Tech Lab has introduced a rival framework known as AAMP, which builds upon existing standards like OpenRTB and OpenDirect, creating a potential split in the market. While some experts view these as separate layers, others point to the overlap in functionality as a source of confusion for developers and marketers alike. Additionally, recent data has indicated that inventory purchased through agentic workflows currently clears at a lower price point compared to traditional programmatic demand, creating a potential revenue gap for publishers. Policy shifts from major tech providers have also added complexity, with some platforms banning the use of certain AI proxies that wrap around their existing advertising capabilities. These factors create a complex landscape where the survival of the protocol depends on its ability to demonstrate tangible value and achieve a critical mass of users.
In light of these developments, market participants adopted several strategic measures to ensure the long-term viability of automated ad buying. Agencies and publishers worked together to establish a unified bridge between the AdCP and AAMP frameworks, recognizing that a fragmented standard would only increase technical debt. They prioritized the implementation of the Trusted Match layer to handle high-speed impression activation, which allowed them to compete more effectively with legacy real-time bidding systems. To mitigate concerns about transparency and fraud, the community shifted toward more rigorous agent authentication methods and expanded the use of the adagents.json registry. Marketers also focused on integrating human oversight into the automated loop, particularly for high-stakes campaigns that required strict adherence to regional compliance laws. By focusing on reducing integration costs and improving the clearing price for agentic demand, the industry moved toward a more equitable and efficient marketplace. These actions established a clear roadmap for future developments, emphasizing that the success of agentic advertising relied not just on technical protocols, but on a shared commitment to transparency, security, and collaborative governance.
