PubMatic launches AgenticOS for agent-to-agent advertising

PubMatic has launched PubMatic AgenticOS, an operating system designed to orchestrate autonomous, agent-to-agent advertising execution across premium digital environments. AgenticOS provides a system-level layer that allows agents to plan, transact, and optimize programmatic advertising with speed, consistency and control. The launch is in close partnership with a group of forward-leaning advertisers, agencies, and publishers actively engaged in testing, shaping, and deploying agent-led workflows, providing real-world validation and innovation for the next phase of AI-native advertising.

Advertising is compounding across new devices and formats, hundreds of thousands of new entrant advertisers, and billions of global users, placing increasing pressure on legacy programmatic systems. In this environment, effective decision-making must happen continuously and coherently, without adding operational complexity or fragmentation. Meeting this opportunity requires advanced systems that can sense, learn, and adapt in real time, while allowing humans greater strategic control and clarity. AgenticOS is designed for this, enabling intelligent execution without latency or operational drag.

The operating system allows advertisers to define objectives, guardrails, brand-safety requirements, and creative parameters in their preferred LLM interface. PubMatic’s platform then operationalizes that intent through a coordinated set of intelligent agents that plan, execute, and optimize campaigns continuously within defined guardrails, allowing teams to focus on higher-value strategy, creativity, and performance analysis rather than manual setup and troubleshooting.

AgenticOS is a synthesis of PubMatic’s multi-year investment in agentic AI and interoperability. From the first agent-to-agent communication specification to yield optimization, audience or inventory discovery and planning, and deal troubleshooting agents, these innovations now come together in AgenticOS. In this unified, end-to-end operating environment, agent-led campaigns are executed, monitored, and continuously refined at scale.

Unlike point execution layers or single-protocol connectors, AgenticOS runs on NVIDIA-accelerated computing and is embedded within PubMatic’s global infrastructure. This foundation enables large-scale, low-latency inferencing and coordination across millions of advertising transactions per second, supporting reliable execution in live markets where performance, scale, and accountability are essential.

In December 2025, PubMatic collaborated with independent agency Butler/Till and Geloso Beverage Group’s Clubtails to launch an early agentic campaign on AgenticOS. Guided by natural-language input through Claude, PubMatic AgenticOS autonomously recommended campaign tactics, executed the buy, and optimized performance in real time within predefined parameters. This automation allowed the agency team to focus on higher-value strategic planning, creative development, and measurement, while the platform handled continuous execution and optimization.

AgenticOS is built on PubMatic’s Architecture of Advertising Intelligence, a three-layer structure designed to support autonomous decision-making at scale while maintaining control and accountability.

At the infrastructure level, the platform uses accelerated computing from NVIDIA to enable microsecond-level inference and real-time, privacy-safe data integration across tens of millions of auctions per second. PubMatic reports that this setup delivers decisioning up to five times faster, sub-millisecond response times, and fewer auction timeouts.

The application layer embeds agentic capabilities across PubMatic’s platform. These agents interpret intent expressed through protocols such as the Ad Context Protocol and Model Context Protocol, and automate core advertising functions including planning, forecasting, pacing, yield management, troubleshooting, and measurement. PubMatic says early tests indicate campaign setup time has been reduced by 87 percent, while issue resolution time has fallen by 70 percent.

The transaction layer connects this agentic decisioning directly to PubMatic’s Activate buying platform. This interoperable execution environment combines premium supply, in-flight campaign signals, and direct bidding infrastructure to transact in real time across Programmatic Guaranteed and Private Marketplace deals. The company positions this layer as a way to keep agent-led workflows compatible with existing buying models while allowing execution to become more adaptive over time.

Amanda Grant, EVP, Global Head of Data and Tech Partnerships, WPP Media: “WPP and Choreograph are using agentic AI to strengthen WPP Open and Open Intelligence. Partnering with PubMatic to test AgenticOS underscores our commitment to agile innovation and to advancing what autonomous advertising can deliver at scale.”

Scott Ensign, Chief Strategy Officer, Butler/Till, a results-driven marketing agency: “We're very excited about the deployment of AI to bring speed and agility into programmatic through agentic media buying. PubMatic has been a critical partner with their AgenticOS in helping us and our clients get to market fast to test and validate the potential benefits of this emerging technology."

www.pubmatic.com

Follow Marketing Report on LinkedIn