AI agents are poised to have a greater role in the CTV ad planning, optimization and selling and process, with a few different developments announced this week amid the Consumer Electronics Show in Las Vegas.
Some introduced agentic AI tools and platforms to help simplify ad campaign planning, optimization and execution, such as Disney, PubMatic and the Viant DSP.
Others like NBCUniversal are also bringing AI agents into the mix for automated selling and buying, as seen by an announcement with FreeWheel, Newton Research and RPA for agent-to-agent buying.
And as AI agents are explored and tested by various companies and platforms, industry body IAB Tech Lab wants to bring some consistency and order to the ecosystem with the release of an Agentic Roadmap that’s meant to help scale agent-to-agent digital ad transactions across the industry.
Here are more details from some of this week’s agentic AI advertising announcements:
Cross-platform agentic media buying powered by NBCU, FreeWheel and Newton Research
NBCU at the beginning of the week announced a partnership to introduce a new approach to buying video advertising using agentic AI across both linear and digital.
It’s powered by and in partnership with NBCU’s SSP FreeWheel, Newton Research and independent agency RPA.
In what NBCU touts as a first-of-its-kind proof-of-concept, the partners released a demo video showing how an agency like RPA can execute and optimize a single premium video investment across linear and streaming inventory in seconds, using NBCU and FreeWheel agents to orchestrate in real-time.
The aim is a quicker, more optimized way to active premium video with reengineered media operations and workflow that take out some of the manual work to make it more efficient.
“This partnership illustrates the potential of agentic AI to hyper-streamline strategic media intelligence and transactions in service of business outcomes,” said Jim Helberg, CEO, RPA, in a statement.
You can watch the demo video here, which shows an end-to-end example of agentic media buying workflow with a premium publisher. It’s powered by AI agent-to-agent collaboration that leverages Model Context Protocol (MCP). FreeWheel and NBCU are deploying AI sales agents for digital and linear respectively, while Newton Research designed and implemented buy-side agents with RPA.
The video explains how, starting with human strategy, advertiser objectives and past campaign insights, the RPA buyer agent initiates a conversation with NBCU’s seller AI agent, telling it about campaign details and goals and to request a buying proposal for linear and streaming.
In the demo, the NBCU AI agent reviews and determines what inventory’s available based on the advertiser objectives, marketplace knowledge and other expertise. It then responds to the buyer agent with a curated set of linear and streaming inventory recommendation options. The agency evaluates the suggestions and selects initial direction based on the goals, and the RPA agent requests more details to refine the plan, after which an NBCU agent delivers a more comprehensive proposal and recommends an optimized linear and streaming investment split. After reviewing, RPA confirms that the recommendations align with goals and moves forward with the buy.
By that point in the process NBCU noted humans are still in the loop as they are today, including to review terms and conditions, creative specs and key execution details. Once approved by the buyer, NBCU automatically creates the order in its order management system and outlines next steps.
The first live execution is set to feature a prominent brand’s Q1 2026 investment, which includes live football playoff games – a factor the partners said mark the first time ever AI agents are automating live sports inventory on linear television.
“NBCUniversal is showcasing that agentic AI is part of the future of media buying, even for premium live sports content. We expect agentic AI collaboration to further automate the way inventory is bought and sold and we are on the leading edge of this revolution with FreeWheel, Newton Research and RPA,” said Mark Marshall Chairman, Global Advertising & Partnerships at NBCU, in a statement.
With the launch of agentic buying, the companies emphasized it’s about creating efficiencies while preserving human expertise. Newton’s agents interoperate and collaborate with other agents, data and tech companies “to create a cohesive intelligence standard for RPA which will ultimately be used for the full campaign planning, execution and measurement loop.”
The use of AI agents is also meant to help put focus and deliver on measuring and optimizing investments, including CTV, for advertiser outcomes.
“From product development to campaign execution, AI has the potential to fuel CTV’s evolution into an automated, outcome-oriented channel. This first-of-its-kind solution is a fundamental step on this journey,” stated Mark McKee, general manager at FreeWheel.
But as AI agents start to play a role in separate ad planning, sales and buying avenues, it can mean more siloed efforts. To that end, IAB Tech Lab wants to inject some consistency and uniformity via standardization to help enable agent-to-agent digital video ad transactions move from experimentation to real-world executions at scale.
IAB Tech Lab releases agentic AI roadmap
On Wednesday IAB Tech Lab unveiled its Agentic Roadmap, which outlines how to take agent-to-agent ad buying from test to scalable execution.
An aspect the organization emphasized is that agentic AI buying won’t require reinventing the ecosystem, but systems can be scaled by extending existing industry standards with new protocols layered in for speed, security and interoperability.
Per IAB Tech Lab, it is one of the first concreate frameworks as to how agentic digital ad buying and selling can operate across the ecosystem, without hindering performance, trust or governance.
And to ensure that the industry body can deliver on the roadmap in a timely manner, IAB Tech Lab said it’s making a “significant engineering investment” focused only on AI development.
“Agentic execution is already part of how digital advertising operates today,” said Anthony Katsur, CEO of IAB Tech Lab, in a statement. “Open, interoperable standards are what make that possible, and our focus is on scaling it responsibly. The fastest and smartest way forward is to build on an existing shared foundation, not introduce multiple new standards that create fragmentation.”
A few key aspects of the announcement:
- IAB’s roadmap builds directly on existing standards like OpenRTB, VAST, measurement, privacy and taxonomies rather than developing new ones
- To support machine speed and agent-to-agent execution modern protocols like MCP, Agent2Agent and gRPC are being integrated
- Plans for 2026 include open source buyer and seller agents, reference implementations, and new trust and measurement signals
- The work is designed to support real adoption – not just tests - by agencies, publishers, brands and ad tech
The organization said industry participants can adopt agentic workflows faster, make integration less complex and lower operational risk by updating the buying execution rather than reinventing or recreating foundational elements.
“The industry gets the best of both worlds,” Katsur continued. “High-performance agentic execution combined with the interoperability, governance, and trust the market already relies on, without reinventing everything.”
NBCU’s Ryan McConville gave an endorsement to the efforts saying, “we believe that interoperability is essential to maintaining performance while unlocking the next phase of automation, and we look forward to continued collaboration.”
Also sharing support was PubMatic’s Nishant Khatri, EVP of Product Management.
“Agentic advertising only scales when intent, execution, and governance evolve together,” stated Khatri. “IAB Tech Lab’s roadmap shows how agentic execution can be integrated into the existing digital advertising foundation, accelerating adoption while preserving interoperability across the ecosystem.”
PubMatic itself is among those that made agentic AI advertising announcements this week.
PubMatic launches AgenticOS
On Monday the ad tech company launched PubMatic AgenticOS, an operating system designed to orchestrate autonomous agent-to-agent advertising execution across premium digital environments.
AgenticOS is a system-level layer embedded within the ad tech player’s global infrastructure that allows agents to plan, transact and optimize programmatic advertising.
Running on NVIDIA computing, PubMatic said that unlike point execution layers, the embedded OS foundation enables large-scale, low-latency inferencing and coordination across millions of advertising transactions per seconds. This supports reliable ad buys in live markets while maintaining performance, scale and accountability.
“We're witnessing the biggest transformation in programmatic since real-time bidding, and agentic media buying is at the center of it,” said Skyler McGill, Head of Video and Programmatic at independent marketing agency Wpromote, a partner trying PubMatic’s operating system. “The shift from theory to practice is happening remarkably fast, launching full campaigns through conversational AI interfaces, optimizing autonomously in real-time.”
Early partners on the effort include WPP Media, Butler/Till, MiQ and others.
“Agentic workflows are poised to reshape how CTV is planned and optimized. PubMatic’s AgenticOS is an important step toward making that future interoperable and scalable across the open internet,” said Jules Minvielle, Cofounder and CEO, Olyzon, an agentic CTV advertising platform. “We’re excited to explore the role agent-led buying can play in delivering greater transparency and performance for advertisers.”