Blab Tech
PubMatic Launches Governance Architecture for Agentic Advertising, Rise Agency Tests It
In a move that could reshape the way brands interact with their audiences, PubMatic announced on 5 May 2026 that it has added a new guardrail architecture to its AgenticOS operating system. The system, described as a “customisable governance framework purpose‑built for autonomous advertising,” is designed to give buyers and agencies full control over how autonomous agents plan, execute and measure campaigns, while producing a complete audit trail.
The first agency to pilot the new framework is Rise, a full‑service media agency that is part of the Quad group. According to a PubMatic press release, Rise will use the architecture to test its performance‑first, transparency‑driven approach to agentic advertising. The company says the framework allows buyers to set rules, thresholds and approval structures that apply to every campaign run through AgenticOS.
AgenticOS was launched in January 2026 as the first operating system built specifically for autonomous, agent‑to‑agent advertising execution across premium digital inventory. The platform already supports more than 80 live campaigns that run on over 100,000 websites, apps and streaming services, and has executed more than 4,000 direct deals. PubMatic reports that advertisers using AgenticOS have seen up to 50 % improvement in total fee efficiency, up to 87 % faster campaign setup times, and up to 70 % faster issue resolution.
The new governance architecture follows a five‑step framework:
1. Define the Boundaries – PubMatic sets universal platform‑wide constraints, while buyers add their own policies such as budget limits, creative approval requirements, inventory allowlists and audience restrictions. 2. Set the Starting State – Before a campaign begins, an agent inherits a “business rules” configuration that encodes the buyer’s targeting, pacing and creative parameters. 3. Arm the Agent – Agents draw from pre‑approved libraries of inventory, creative and audience segments that have already cleared the buyer’s review process. 4. Keep Humans in the Decision – If an agent encounters a decision that exceeds its authority, it escalates to an approver who authenticates via a one‑time passcode, creating a verified record of the intervention. 5. See Everything, Always – Every action is logged, and drift detection monitors behavior against expected parameters, flagging deviations for human review.
The architecture is embedded at the point of execution, inside the same infrastructure where campaigns are built and activated. This contrasts with post‑hoc governance solutions that are layered on after the fact.
Klaudia Smykowska, Group Director of Media Investment at Rise, said the framework “provides the verifiable control and audit trail that buyers need to trust autonomous advertising.” She added that “transparency and accountability have always been non‑negotiable for us. They’re core to how we work and how we earn our clients’ trust.”
Nishant Khatri, EVP of Product Management at PubMatic, explained that the architecture was built after extensive discovery work with buyers. He said the five‑step model “answers what buyers told us they needed: controls, visibility and a foundation that can scale without a ceiling.”
The guardrail architecture is part of a broader set of accountability tools in AgenticOS, including a Fee Transparency Agent and a Detailed Reasoning Agent that explain each dollar in the supply chain and the system’s decision logic.
As the industry moves toward more autonomous buying, the ability to audit and control agentic decisions is expected to become a key differentiator. PubMatic’s architecture, with its embedded guardrails and real‑time approval workflows, aims to provide that level of trust.
At present, Rise is the only agency publicly confirmed to be piloting the framework. PubMatic has not announced additional agency partners, but the company says the architecture is “configurable once and works everywhere.” The next steps for the industry will likely involve broader adoption, regulatory scrutiny of autonomous ad tech, and further performance data from live campaigns.
The launch of PubMatic’s governance architecture marks a significant step toward making autonomous advertising more transparent and accountable, potentially setting a new standard for how agencies and buyers manage AI‑driven campaigns.
The first agency to pilot the new framework is Rise, a full‑service media agency that is part of the Quad group. According to a PubMatic press release, Rise will use the architecture to test its performance‑first, transparency‑driven approach to agentic advertising. The company says the framework allows buyers to set rules, thresholds and approval structures that apply to every campaign run through AgenticOS.
AgenticOS was launched in January 2026 as the first operating system built specifically for autonomous, agent‑to‑agent advertising execution across premium digital inventory. The platform already supports more than 80 live campaigns that run on over 100,000 websites, apps and streaming services, and has executed more than 4,000 direct deals. PubMatic reports that advertisers using AgenticOS have seen up to 50 % improvement in total fee efficiency, up to 87 % faster campaign setup times, and up to 70 % faster issue resolution.
The new governance architecture follows a five‑step framework:
1. Define the Boundaries – PubMatic sets universal platform‑wide constraints, while buyers add their own policies such as budget limits, creative approval requirements, inventory allowlists and audience restrictions. 2. Set the Starting State – Before a campaign begins, an agent inherits a “business rules” configuration that encodes the buyer’s targeting, pacing and creative parameters. 3. Arm the Agent – Agents draw from pre‑approved libraries of inventory, creative and audience segments that have already cleared the buyer’s review process. 4. Keep Humans in the Decision – If an agent encounters a decision that exceeds its authority, it escalates to an approver who authenticates via a one‑time passcode, creating a verified record of the intervention. 5. See Everything, Always – Every action is logged, and drift detection monitors behavior against expected parameters, flagging deviations for human review.
The architecture is embedded at the point of execution, inside the same infrastructure where campaigns are built and activated. This contrasts with post‑hoc governance solutions that are layered on after the fact.
Klaudia Smykowska, Group Director of Media Investment at Rise, said the framework “provides the verifiable control and audit trail that buyers need to trust autonomous advertising.” She added that “transparency and accountability have always been non‑negotiable for us. They’re core to how we work and how we earn our clients’ trust.”
Nishant Khatri, EVP of Product Management at PubMatic, explained that the architecture was built after extensive discovery work with buyers. He said the five‑step model “answers what buyers told us they needed: controls, visibility and a foundation that can scale without a ceiling.”
The guardrail architecture is part of a broader set of accountability tools in AgenticOS, including a Fee Transparency Agent and a Detailed Reasoning Agent that explain each dollar in the supply chain and the system’s decision logic.
As the industry moves toward more autonomous buying, the ability to audit and control agentic decisions is expected to become a key differentiator. PubMatic’s architecture, with its embedded guardrails and real‑time approval workflows, aims to provide that level of trust.
At present, Rise is the only agency publicly confirmed to be piloting the framework. PubMatic has not announced additional agency partners, but the company says the architecture is “configurable once and works everywhere.” The next steps for the industry will likely involve broader adoption, regulatory scrutiny of autonomous ad tech, and further performance data from live campaigns.
The launch of PubMatic’s governance architecture marks a significant step toward making autonomous advertising more transparent and accountable, potentially setting a new standard for how agencies and buyers manage AI‑driven campaigns.
