In Liu Cixin’s novel, there’s a scene where thousands of people are arranged like a living supercomputer – each person acting as a tiny logic gate, trying to solve a calculation that a real machine could finish in an instant. That’s not far from how today’s programmatic ad auctions work. They’re incredibly complex, running hundreds of millions of times a day, and every single one needs to decide – within a fraction of a second – whether to show your ad, to whom, and at what price.
The problem is that those decisions can involve multiple systems communicating under tight time constraints. For most people, milliseconds sound fast. In programmatic advertising, however, even small delays can affect how much data and how many decisioning steps fit into the available auction window.
The IAB Tech Lab has introduced a framework designed to address some of these infrastructure challenges. And no, this isn't another buzzword bingo situation. This one actually matters.
What Just Happened?
On November 13, 2025, the IAB Tech Lab released the Agentic Real Time Framework (ARTF) v1.0 for public comment. They introduced a new standardized approach to interoperability across the programmatic advertising ecosystem through the use of containerized architecture. Think of it as a rethinking of how the pipes of programmatic advertising actually work.
Since then, ARTF v1.0 has been finalized. The final specification is available through IAB Tech Lab and the official GitHub repository. The framework defines how agent services can run as containers within a host platform and be called directly by that platform to perform specific tasks.
Here's the simple version: today's programmatic ecosystem is like having a conversation where every sentence has to travel through different buildings, different cities, sometimes different cloud environments. Your DSP talks to an SSP, which might need to check with a data vendor, which pings a fraud detection service, and so on. All of that back-and-forth takes time – precious milliseconds that add up fast.
The new framework is designed to reduce this overhead. By allowing approved agent services to run within the host platform's own infrastructure, ARTF can reduce the need for repeated communication between separate systems over external network connections. IAB Tech Lab has described a potential reduction of up to 80% in bid request and response time. This is a stated performance target, not a guarantee for every implementation: actual latency depends on the architecture, workload, integrations, and deployment environment.
How? By letting different systems – such as DSPs, SSPs, fraud detection services, and data enrichment tools – run containerized services within the same host environment. Instead of sending every request to a separate external service, a host platform can call an approved service running in its own infrastructure. This can reduce communication overhead while allowing the service provider's code to remain isolated.
Why You Should Care
Let's cut through the technical jargon. Here's what this means in practical terms:
Speed That Changes Everything
When auction processing becomes faster, something interesting happens. You're not just making things faster – you may create more room for operations that need to happen before a bid decision is made. More sophisticated fraud detection before you bid on an impression, richer data analysis in real time, and additional opportunities for decisioning could become more practical within a limited auction window.
Remember Liu Cixin's human computer? The genius wasn't in the individual people – it was in how they communicated. ARTF is about making those connections more efficient instead of having each "logic gate" shout across a crowded room.
The extent of these benefits depends on the implementation. Reduced communication overhead may create more room for real-time processing, but it does not automatically guarantee better bidding decisions, improved fraud detection, or higher campaign performance.
The Agent Question
You've probably heard "AI agents" thrown around a lot lately. The ARTF framework is building infrastructure to support agentic advertising by defining how agent services can be deployed and called within a host platform.
ARTF is specifically a framework for real-time agent services in advertising infrastructure. It should not be confused with a general-purpose AI assistant or with protocols designed for broader buyer–seller interactions. Its focus is on how services can run within a host environment and participate in real-time bidstream processing.
Security Without Sacrifice
Here's something clever: containerization can help isolate code and control how services interact with the host platform. A DSP or SSP can run an approved service without necessarily exposing all of its internal systems or proprietary code to the service provider.
ARTF defines requirements intended to support protected and private bidstream processing. The framework's container model is designed to keep agent services isolated within the host's infrastructure, with external network access restricted by default. Actual security still depends on correct implementation, configuration, and operational controls.
Industry Alignment
This isn't just a theoretical experiment. The framework has been developed through IAB Tech Lab's industry work, with participation from companies across the programmatic advertising ecosystem.
IAB Tech Lab has announced that ARTF v1.0 is already being used in production by multiple companies, including organizations involved in its development. This shows that the framework has moved beyond the public-comment stage, although adoption and capabilities vary by implementation.
What's in the Framework?
The ARTF introduces several key technical components:
- Containerized Architecture: Instead of requiring every service to communicate across separate network infrastructure, approved services can run in containers within the host platform's environment, potentially reducing communication overhead.
- Agent-to-Agent Communication: The framework supports agent-driven services and interfaces, including Model Context Protocol (MCP) integration. This can help host platforms interact with specialized services through defined interfaces.
- Better Security: Containers can be isolated and governed through defined runtime requirements and controls. This helps platforms manage how external services run within their infrastructure.
- Efficient Real-Time Processing: By reducing unnecessary network communication, ARTF is designed to help support latency-sensitive operations within the programmatic bidstream. The actual performance and resource-efficiency benefits depend on the deployment and workload.
ARTF v1.0: From Public Draft to Finalized Specification
ARTF v1.0 is no longer awaiting public comments or finalization. Following the public-comment period, IAB Tech Lab published the finalized specification, establishing a stable foundation for implementations of containerized agent services in real-time advertising infrastructure.
The specification defines standard requirements for container runtime behavior and an API for reliable, protected bidstream modification. It also describes how agent services can declare their capabilities and how a host platform can manage and invoke those services.
The framework is intended to let service providers package their capabilities once and deploy them to compatible host platforms, while the host manages the operational environment. This could make it easier to integrate specialized services into real-time workflows without requiring every service to operate across separate infrastructure.
The finalized v1.0 specification is available in the IAB Tech Lab documentation and the official GitHub repository. Work on a future version is also underway, as the industry continues to explore additional use cases.
Where the Industry Stands
ARTF is an industry framework, and its publication does not mean that every DSP or SSP has implemented it. Adoption depends on individual platforms deciding to support the specification and integrating compatible services into their infrastructure.
The framework should also be distinguished from other agentic advertising initiatives, such as the Ad Context Protocol (AdCP), which focuses on interactions between advertising buyer and seller systems. ARTF addresses a different layer: deploying and invoking agent services within a host platform's real-time infrastructure. Support for one initiative does not establish support for the other.
The Bigger Picture
Step back for a moment. ARTF isn't only about making auctions faster (though that's nice). It's about building infrastructure for a more connected programmatic advertising ecosystem – one where specialized services can participate in real-time decisioning without every operation requiring communication across separate systems.
As industry discussions around ARTF have highlighted, this kind of efficiency has historically been easier to achieve within tightly controlled, single-platform environments. ARTF aims to make similar approaches more accessible to the open programmatic ecosystem.
In "The Three-Body Problem," the human computer eventually fails – not because the people weren't smart enough, but because human communication is too slow. ARTF addresses a version of that challenge in programmatic advertising: helping existing systems and specialized services communicate with less overhead.
Will it work? The finalized specification and reported production deployments show that ARTF has progressed beyond an early concept. The scale of its impact will depend on adoption, implementation quality, and the use cases being deployed. The technical foundations – containerization, standardized service interfaces, and security controls – provide a framework for further development in programmatic advertising.
What Should You Do?
So what does this mean for you? Programmatic advertising continues to evolve, and infrastructure improvements may create new opportunities for real-time processing. Advertisers should still focus on what matters: knowing where their ads run, what they're paying, and whether their campaigns are working.
For advertisers and publishers, the practical takeaway is to understand ARTF as an industry framework that could help reduce communication overhead in real-time advertising systems. The stated reduction of up to 80% should be treated as a potential outcome, not a universal performance guarantee.
Want to see how our platform performs today, not in some theoretical future? Create an account on SmartyAds DSP, and we'll show you how the platform works and whether it fits what you need.
