On September 22, 2026, the IAB Tech Lab released version 3.0 of its Agentic Advertising Management Protocols (AAMP). This update shifts focus from ad execution to campaign planning. The release follows version 2.3 by less than two months.
The protocol introduces OpenProposal, a specification that automates how media buyers and sellers discover and evaluate advertising inventory. Buyer agents use this standardized format to read publisher offerings. The system matches campaign briefs with available ad products without manual intervention.
Automating Media Planning
Previous AAMP iterations focused on execution, such as standardizing data schemas for AI tools to interpret creatives or content categories. OpenProposal extends this architecture into the initial planning phase.
Publishers package ad products into machine-readable formats. This structure helps smaller sellers gain visibility because buyer agents automatically evaluate their inventory. For advertisers and agencies, automating discovery reduces manual research. Consequently, human teams focus on strategy rather than administrative data matching.
OpenProposal allows buyer agents to evaluate a wide range of opportunities, helping them to create more diverse campaigns and freeing up more time on the human side for strategy.
Technical Standardization Across the Supply Chain
The AAMP framework adapts existing industry schemas to ensure independent AI agents operate with a shared vocabulary. When different AI systems use identical taxonomies, they maintain a consistent understanding of content compliance and audience definitions.
After an AI agent confirms a media plan through OpenProposal, the AAMP framework translates that plan into execution. This pipeline connects the discovery phase directly to programmatic ad delivery.
IAB UK announced that OpenProposal has entered a public comment period until October 22, 2026. The IAB Tech Lab seeks feedback from industry stakeholders to refine the specification before final implementation.
Source: IAB UK News
This article was created with AI assistance and editorially reviewed.
