AAMP vs AdCP: What IAB Tech Lab and AAO Are Each Building for Agentic Advertising

Agentic advertising now has two standards bodies

Eighteen months ago, agent-driven ad buying was a keynote topic. Today it is two competing standards stacks with shipped software. IAB Tech Lab released AAMP 2.0 in June 2026, adding transaction-ready buyer and seller agent SDKs to the framework it launched in January. AgenticAdvertising.Org, the trade association formed around the Ad Context Protocol, shipped AdCP 3.0 in April and already has 3.1 release candidates in review.

The two efforts are not coordinated. Tech Lab publicly positioned AAMP as the standards-grounded alternative to AdCP, and AAO kept building anyway. Programmatic teams do not get to wait out the fight: buyer agents, seller agents, and the platforms between them are adopting pieces of both, and every video transaction in either stack still resolves to the same artifact at the delivery layer, a VAST tag.

The IAB Tech Lab stack: AAMP

AAMP (Agentic Advertising Management Protocols) is Tech Lab's umbrella for everything agentic. It rests on three pillars. Agent Foundations contains ARTF, the Agentic Real Time Framework, which defines how agent services run as containers inside host platforms, including real-time bidders. Agentic Protocols covers the Buyer and Seller Agent SDKs plus Agentic Direct, Agentic Deals, and Agentic RTB. Trust and Transparency is anchored by an Agent Registry.

The design philosophy is explicit: build on the rails that already move money. AAMP wires OpenRTB, AdCOM, OpenDirect, and VAST together, with the Model Context Protocol and Google's Agent2Agent protocol as the agent interfaces.

Version 1.0 could only transact direct deals. The 2.0 release added programmatic. The Buyer Agent SDK now creates and optimizes programmatic deals through a three-layer hierarchy of orchestration agents, channel specialists, and functional agents. The Seller Agent SDK turns a static media kit into a storefront that adapts pricing and packaging to the buyer and the deal stage. A Deals Library acts as the system of record, with native OpenDirect 2.1 and Deals API support. Approval gates and audit logs keep humans at the sign-off points.

The AAO stack: AdCP

The Ad Context Protocol launched in October 2025 from a consortium that included Yahoo, PubMatic, Optable, Scope3, Swivel, and Triton Digital. Governance has since been formalized under AgenticAdvertising.Org (AAO), a trade association with four equally weighted voting classes: brands, agencies, publishers, and technology providers. The reference sell-side implementation lives with the Prebid community, deliberately separate from spec governance.

Where AAMP extends existing rails, AdCP was designed agent-native from the start. It is built directly on MCP and runs asynchronously, so humans can approve while agents negotiate. AdCP 3.0, released in April 2026, stretched the protocol across the full campaign lifecycle: inventory discovery, media buys, creative production with brand.json and the Creative Protocol, governance, and reporting across 20 media channels. The working group is already iterating on 3.1, which tightens the creative lifecycle with retention contracts and status webhooks.

Adoption is real on the sell side. AdCP is in production at Snap, Pinterest, Reddit, Netflix, and Vox Media.

Where the stacks actually differ

• Governance: IAB Tech Lab is the incumbent body that already maintains VAST, OpenRTB, and OpenDirect. AAO is a new trade association built specifically for agentic advertising, with brands and agencies holding equal votes to publishers and vendors.
• Design center: AAMP layers agents on top of the existing programmatic rails. AdCP defines new MCP-native tasks first and treats legacy systems as integration targets.
• Coverage: AAMP is deepest at the transaction layer, with ARTF containers running inside bidder infrastructure. AdCP is broadest across the lifecycle, from planning through creative production to reporting.
• Interfaces: both name MCP as a core agent protocol, which means a tool exposed as an MCP server can serve agents on either stack without modification.
• Verification: neither stack validates the creative payload. Both move deal state and context between agents; the correctness of what actually ships is out of scope for both specs.

Both stacks still hand off to VAST

Follow a video buy through either stack and the endpoint is identical. An AAMP agentic deal or an AdCP media buy ultimately serves through an ad server that returns VAST XML: an InLine or Wrapper, MediaFiles, Impression pixels, TrackingEvents. AAMP names VAST among the standards it builds on. An AdCP creative agent that produces video hands a VAST tag to the seller side. Twenty channels of protocol surface on one side, one XML document on the other.

That document is where delivery fails. A missing Duration, an HTTP media URL on a CTV device that requires HTTPS, a wrapper chain that exceeds the player's depth limit, a VPAID creative on an inventory path that dropped VPAID years ago: none of it is visible at the protocol layer. The deal confirms cleanly, the agents log success, and the impression dies at runtime with a VAST error code nobody is watching.

Human traffickers used to be the backstop. In an agent-to-agent transaction there is no trafficker. If the tag gets checked at all, something in the loop has to check it, which means validation has to be a tool the agent calls before the deal confirms, not a QA pass scheduled after launch.

What to do while the standards fight plays out

Betting on a single winner is unnecessary. MCP is the shared interface: AAMP's Buyer Agent SDK accepts additional MCP tool servers, and AdCP is MCP-native end to end. A validator exposed over MCP plugs into both stacks today.

vastlint runs 187 rules grounded in the IAB VAST, OMID, and SIMID specs and returns deterministic, machine-readable results. It is available as a hosted MCP endpoint at vastlint.org/mcp, as a CLI for pipelines, and as native libraries. Point either stack's agent at it and gate deal confirmation on a clean tag. The protocols will keep changing; the payload they deliver has been VAST for two decades and will still be VAST when the governance questions settle.