The Future of Ad Tech: CNN International’s Bold Bet on Agentic Workflow

By Seb Joseph
October 22, 2026

As the media landscape stabilizes following the recent Skydance acquisition of Paramount and the broader restructuring within the Warner Bros. Discovery (WBD) ecosystem, one question looms large over the industry: How will advertising evolve when the human middleman is replaced by autonomous software?

For CNN International, the answer is already being written. Despite the seismic shifts at the corporate level, the network’s ambitious push into "agentic" advertising—a system where AI agents negotiate, book, and optimize ad campaigns on behalf of brands—remains firmly on track. Seven months after first announcing its intent to pivot toward this automated future, the network confirms it is still aiming for a Q1 2027 launch.

The Architecture of Automation: Main Facts

At its core, CNN International is developing a proprietary, in-house agentic infrastructure designed to interface with third-party demand-side platforms (DSPs) and buyer-side agents. The goal is to move beyond the manual "RFP-to-trafficking" grind that has defined media buying for decades.

Faisal Karmali, Vice President of Digital Operations for CNN International Commercial and WBD, remains optimistic about the timeline. "I gave a Q1 timeline to have the first dollars flowing, which I would still be hopeful for," Karmali stated during the recent Digiday Media Strategies event in London.

However, the broadcaster is playing a cautious, strategic game. Rather than rushing to market, the team is prioritizing the construction of a robust, scalable system that can withstand the complexities of modern digital advertising. This involves teaching AI agents the "business rules" of CNN—everything from brand safety standards to inventory pricing—and ensuring that the network’s ad inventory is machine-readable for the large language models (LLMs) that power buyer-side agents.

Chronology of Development

The path to this launch has been methodical, adhering to a structured development timeline:

• Early 2026: CNN International announces its intent to build in-house agentic infrastructure, signaling a move toward autonomous ad trading.
• Q2 2026: The network focused on scoping technical protocols and establishing the foundational rules for how an AI agent should interact with their inventory.
• Q3 2026: Testing phases commenced, specifically observing how one or two key properties were interpreted by LLMs. This period was critical for verifying that metadata layers were correctly identified by external systems.
• Q4 2026 (Current): The focus has shifted toward analyzing buyer behavior and budget allocation. The network is now deep in "conversational" phases with technology partners, attempting to iron out the friction points of automated trading.
• Q1 2027 (Projected): The official launch of the agentic workflow, with initial use cases centered on performance-driven campaigns before expanding into more complex brand-building efforts.

Supporting Data and Technical Hurdles

While the promise of efficiency is high, the technical execution is a feat of engineering. The primary challenge, according to industry observers, is that the industry is currently building in silos. Buyers are developing their own proprietary agents, while publishers like CNN are building their own, creating a "Jenga-like" scenario where every adjustment to a protocol can cause instability elsewhere.

"I feel like both sides need to talk while they’re building their platforms," said Karmali. "At the moment, everything gets built in silos, and then we have to figure out how to connect it in the middle."

The "Language" Barrier

The industry lacks a universal standard for how these agents communicate. There are currently competing frameworks, and there is no agreed-upon mechanism to prevent agents from "gaming" the system—such as misinterpreting a brief to prioritize cheaper, lower-quality inventory.

Furthermore, there is the issue of "provenance." CNN relies on its reputation for high-quality, fact-checked journalism. However, an AI agent on the buy side doesn’t "see" a brand’s reputation; it only sees the metadata. If the metadata is not perfectly aligned with what the buyer’s agent is looking for, the inventory becomes invisible.

"From an agent perspective, in terms of exposure of the layer beyond MCP (Media Consumption Protocol) and looking at a common language for us to trade with agencies, DSPs and clients directly, I think that’s the important bit," Karmali noted.

Expert Analysis and Official Perspectives

The industry is divided on whether agentic workflows will truly revolutionize performance or merely automate the paperwork.

Ravi Patel, CEO of SWYM, an ad-tech startup specializing in programmatic decisioning, offers a pragmatic warning: "Seller agents are a genuine efficiency gain for direct sales. They take cost out of the RFP-to-trafficking grind. But they automate the deal, not the delivery. Performance is decided impression by impression, in milliseconds, on bid-request signals. Connect those two layers and you’ve got something. Otherwise, you’ve automated the paperwork, not the performance."

CNN appears to acknowledge this distinction. By focusing on "clean metadata" and "human-made, fact-checked content," the network is betting that their premium environment will be a prioritized destination for buyer agents once the technical standards are unified.

Karmali refused to speculate on whether this shift would fundamentally rebalance power in favor of publishers, noting that it is "too early to say." When asked about the potential revenue impact, he provided a broad range: agent-led deals could start at 5% of CNN International’s ad business, but there is a long-term potential to reach 50%.

Strategic Implications for the Media Ecosystem

CNN International’s strategy is emblematic of a broader trend: traditional media houses are diversifying their revenue streams while simultaneously "future-proofing" their operations.

Beyond the AI experiment, the network is aggressively pursuing other growth vectors:

• FAST Channels: Their Free Ad-supported Streaming TV (FAST) channels now operate in 50 countries, leveraging existing digital clip archives to capture new audiences.
• Subscription Models: The CNN All Access subscription service, launched in the U.K. and Canada just days ago, integrates ads from the ground up, proving that even premium, paid-for products are leaning into hybrid monetization.
• Experiential Media: The "Global Perspectives" events series represents the opposite of automation. It is high-touch, human-centric, and sponsor-funded—a hedge against a world where digital trading becomes fully commoditized.

The Road Ahead

As the industry moves toward 2027, the success of CNN’s agentic project will depend less on the AI itself and more on the collaborative spirit of the industry. Without a unified standard for communication, guardrails for trading, and clear definitions of "quality inventory," the project risks falling into the same traps that plagued early programmatic advertising: fraud, lack of transparency, and wasted spend.

For now, CNN International remains a bellwether. By investing in the infrastructure today, they are positioning themselves to be the "first mover" in a market that will eventually demand automated efficiency. Whether that demand leads to a new golden age of trading or a new set of digital headaches remains the defining question for the next year of media innovation.

"There’s still a path to go," Karmali concluded. And for the rest of the industry, that path is one worth watching closely.