In the last few months, I have come across multiple articles, social media posts and event fireside chats that completely miss the mark on the scope of the IAB Tech Lab’s Agentic Ad Management Protocols, also known as AAMP. Usually, they all sound like a straw man designed around presumed roles of AAMP and AdCP.
Rather than looking under the hood of publicly available AAMP (Agentic Advertising Management Protocols) materials, the comparisons seem to start with a wishful hypothetical architecture, thereby missing the clearly stated goals of AAMP.
AAMP and AdCP are agentic advertising frameworks with different design philosophies – not two cleanly separated layers that fit together naturally.
AAMP is designed for continuity: it builds on existing standards and operational concepts that the ecosystem and platforms already understand and use to transact billions of dollars in advertising spend today with seamless integrations into existing use cases and systems.
Most common analytical mistakes in the comparisons can be summed up as:
- Treating ARTF as the agentic standard, rather than just one real-time component of AAMP
- Ignoring l overlap between the two in the majority of business functions supported: planning, discovery, negotiation, ordering, governance, audiences, and campaign execution.
- Describing some sort of hypothetical wishful-thinking hybrid architecture without any basis or logical reasoning in what is actually in the AAMP framework.
- Contradicting themselves. For e.g., first describe AAMP’s Buyer Agent as covering planning, discovery, negotiation, deal creation and campaign automation, but later stating that AAMP solely operates inside the auction platform.
- Misunderstanding the role of MCP and clubbing it along with AAMP and AdCP as if they perform the same functions, Shared MCP support does not make the frameworks complementary. MCP is a communication substrate; it does not reconcile their advertising semantics.
Lets be crystal clear on one thing:
AAMP is not an “impression-layer protocol.” It is an umbrella framework that includes campaign planning, media and audience discovery, transaction management, as well as impression-time execution, demonstrating the breadth of coverage Tech Lab standards provide across the full advertising workflows from campaign planning to execution.
Here is a more detailed inventory of some of the factual errors encountered in these published comparisons:
| Claim | Assessment | Why |
| AAMP extends the impression layer; AdCP defines the campaign layer. | Incorrect | AAMP explicitly includes a management layer for audience and media discovery, planning, negotiation, orders, deals, signals, and pre-execution setup. |
| AdCP and AAMP are not substitutes. | This is too absolute | They are not complete one-for-one substitutes, but major portions are directly substitutable or overlapping. |
| AdCP runs before and after the auction; AAMP runs inside the platform where the auction clears. | Incorrect | AAMP supports the entire transaction, end-to,end. It supports direct transactions, Programmatic Guaranteed, PMPs, deal creation, bookings, and campaign planning. Not merely auction execution. |
| One AdCP media buy produces thousands of AAMP impression events. | Hypothetical, not standardized | No normative AdCP-to-AAMP decomposition or interoperability contract establishes this relationship. |
| ARTF is the nearest counterpart to AdCP Trusted Match. | Weak comparison | ARTF is an extremely low-latency, host-executed agent-service framework supporting several use cases. Trusted Match is only a specific targeting plane. |
| IAB Tech Lab standards carry no view of the CTV pod | Incorrect | OpenRTB 2.6 introduced pod-specific fields such as poddur, podid, podseq, slotinpod, maxseq, and mincpmpersec. AAMP uses existing standards to define its object schema, so a detailed pod definition will always be part of it |
| AAMP primitives assume auction dynamics. | Incorrect generalization | ARTF/OpenRTB address auctions, but AAMP Buyer/Seller Agents support media buying and selling processes and have nothing to do with auctions. |
Claim: Mistake one AAMP pillar for the entire framework.
This is the central category error in the comparisons.
AAMP is made up of three pillars:
- Agentic Foundations, one of which is ARTF.
- Agentic Protocols & Standards, codified in the Buyer and Seller Agent SDKs.
- Agent Registry for Trust and Transparency.
IAB Tech Lab’s own description says AAMP’s Agentic Protocols layer is the management layer and covers how agents:
- Discover buyers/sellers and inventory;
- Discover audiences and unique ad products;
- Understand campaign requirements;
- Negotiate pricing;
- Manage orders and deals;
- Exchange audience signals;
- Complete campaign setup before execution.
That is plainly campaign and transaction lifecycle functionality, not merely impression-layer auction activity. IAB Tech Lab’s AAMP overview.
The AAMP Buyer Agent makes the contradiction even clearer. It accepts a campaign brief, allocates budgets across channels, researches multiple sellers, negotiates prices, obtains human approval, books deals, and produces an audit trail. Its documentation includes linear-TV buying, change requests, pacing, creative management, reporting, and multi-seller orchestration. Official AAMP Buyer Agent documentation.
AAMP is not merely an impression or bidstream layer. Rather, that is ARTF’s role;
AAMP as a whole addresses much more. AAMP’s Agentic Protocols, Buyer Agent SDK, Seller Agent SDK, Agentic Direct, and Agentic Deals explicitly address campaign planning, inventory and audience discovery, negotiation, ordering & booking options, and management.
Claim: AAMP and AdCP have no direct overlap
Most comparisons begin and end with applying a supposed layer distinction. The opposite is closer to the truth.
Both frameworks address:
- Buyer and Seller agents;
- natural-language campaign briefs;
- inventory and product discovery;
- seller discovery;
- media planning;
- pricing and negotiation;
- deal or order creation;
- human approval before spend;
- audience discovery and activation;
- agent identity and registry functions;
- MCP and A2A communication;
- campaign lifecycle and reporting;
- real-time or serve-time decisioning.
The object names and abstractions differ, but the business functions overlap.
Therefore, the comparisons cannot assign those functions exclusively to AdCP and then use that assignment to prove complementarity.
Claim: CTV analysis or other channel analysis almost always contains a direct factual error
The statements such as an RTB bid request providing “no view of the pod” are wrong. CTV is just one example but is repeated for channels too.
OpenRTB 2.6 explicitly introduced ad-pod support for CTV and audio, including:
- poddur
- podid
- podseq
- slotinpod
- maxseq
- mincpmpersec
- exact-duration support
- pod deduplication signaling
Some even declare that “CTV is not genuinely impression-level.” That is rhetoric rather than a useful technical proposition. CTV can be traded at multiple grains:
- campaign or deal;
- show, collection or audience package;
- ad break or pod;
- slot;
- delivered impressions.
A system may need several of those grains simultaneously. That does not make the impression layer unreal.
Treatment of media channels almost always understates AAMP
The arguments that AAMP is structurally disadvantaged for broadcast, CTV, radio, and DOOH because these channels are not always impression-auction environments are uninformed.
AAMP does not require every transaction to clear through OpenRTB. Its architecture includes:
- OpenDirect-based direct transactions;
- Deals API integration;
- PMPs;
- programmatic guaranteed transactions;
- buyer and seller booking workflows;
- linear-TV scatter buying in the current Buyer Agent.
The AAMP Buyer Agent currently supports DMA targeting, CPP and CPM pricing, dayparts, bookings, order state machines, and post-deal changes for linear TV. That directly contradicts the suggestion that AAMP can only function where impression auctions exist.
So, what Is AAMP?
AAMP is an IAB Tech Lab umbrella initiative for bringing agentic AI into the advertising ecosystem. Its design priorities are:
- reuse of established advertising semantics;
- compatibility with deployed infrastructure;
- buyer/seller reference implementations;
- direct, deal-based and programmatic workflows;
- deterministic execution at advertising-system latency;
- registries, diligence, and trust;
- post-campaign reconciliation and billing support;
- incremental industry adoption.
AAMP’s principal advantage is continuity: AAMP’s objective is to build the foundations for deploying AI technology across all forms and shapes of advertising covering end-to-end ad decisioning and delivery workflows.
Shailley Singh
COO and EVP, Product
IAB Tech Lab