IAB Tech Lab introduces ARTF and Agentic Audiences for sub-100ms bidding
The IAB Tech Lab has introduced the Agentic Real Time Framework (ARTF) and Agentic Audiences protocol to facilitate containerized bidding in advertising. These standards aim to enable AI-based contextual inference directly within the browser or server-side auction environment, potentially reducing latency by 80%.
Key Takeaways
- ARTF facilitates containerized bidding, allowing third-party 'agent services' to run directly inside the host auction infrastructure via gRPC.
- Agentic Audiences, formerly LiveRamp’s User Context Protocol, uses dense vector embeddings (256-1024 dimensions) to compress identity and contextual signals.
- Standardization aims to reduce average bid request and response times from several hundred milliseconds to approximately 100 milliseconds.
- The framework supports 'in-loop inference,' where AI models evaluate real-time signals like physical location and social environment during the auction.
Why It Matters
The transition to containerized bidding represents a structural realignment of the open web to compete with walled gardens like Google and Meta, which already co-locate data and compute. By enabling real-time AI inference at the auction level, streaming platforms can move beyond blunt genre-based targeting to hyper-local, situational relevance without the latency or privacy risks of external data calls. For the broader ecosystem, this reduces the 'cloud tax' and data leakage concerns inherent in identity-based targeting. Watch for adoption rates among supply-side platforms (SSPs) using ARTF version 1.0 to enrich bid streams with seat-level contextual metadata.
Additional Context
The development of ARTF follows the IAB Tech Lab’s 2025 roadmap, which prioritized rebuilding signal for publishers as third-party cookies and household IDs lose efficacy. Per IAB Tech Lab (November 2025), the framework is designed to move programmatic advertising toward an 'agentic' foundation where autonomous buyer and seller agents manage millions of decisions per second. This shift is particularly critical for live streaming and CTV, where high concurrency and short ad windows demand sub-millisecond execution. In August 2025, a pilot program involving Zillow, Chalice, and Index Exchange validated the approach by embedding DSP intelligence directly into SSP infrastructure to access granular site-level data. This containerized model addresses longstanding industry complaints about the technical limitations of Google's Privacy Sandbox. Per The Current (February 2024), the IAB Tech Lab previously identified 44 use case gaps in the Privacy Sandbox, specifically regarding latency and frequency capping. By standardizing OCI-compliant containers manageable via Kubernetes, the ARTF allows specialized tech providers to deploy services like fraud detection or bid enrichment across any standard-compliant platform without unique integrations. As of early 2026, the IAB Tech Lab has expanded this initiative into version 2.0 and introduced an Agent Registry to ensure transparency in how these autonomous systems operate across the supply chain.
Read full article at adexchanger.com
Enjoy our coverage?
Add StreamingMeme as a preferred source on Google to see more of our streaming news at the top of your Search results.
Add as preferred source