Fox Automated TV Ad Deals: How AI Agents Cut the Buying Process from Days to Seconds

Buying advertising time on television has always been one of the most paperwork-heavy processes in the media business: hundreds of phone calls, faxes, endless spreadsheets, and email chains stretching across days. Fox, one of the largest players in the media market, tackled this head-on. After deploying AI agents inside its Fox AdStudio platform, the full cycle from campaign planning to activation compressed from days and weeks to seconds. The routine that was burning out planners and sellers moved to agents, and the people went back to strategy and audience work.

The media business is built in a way where every ad-buying deal passes through dozens of manual touchpoints: a call, an email, a spreadsheet, a correction, another email. Multiply that by hundreds of deals a month and you get thousands of hours that skilled specialists spend shuffling paper instead of doing strategy. Every delay in approval is a risk of losing a client or missing a valuable ad slot. That load can be lifted entirely today.

What the TV Ad Buying Process Looked Like Before AI Agents

The traditional TV advertising market is an industry with deeply entrenched manual processes. A planner on the advertiser side and a seller on the media company side could spend weeks exchanging briefs, proposals, and revisions before a deal closed. Not a single stage was truly automated.

In practice, this meant: negotiations through dozens of phone calls and email threads to agree on terms, rates, and formats; document exchange via Excel spreadsheets, PDF media plans, and faxes that had to be received, printed, checked, and returned with corrections; manual forecasting where audience targeting, reach estimates, and placement recommendations all required significant time; and campaign activation that demanded separate manual steps and constant monitoring even after a deal was signed.

The result was that specialists hired for analytics and strategy spent a large part of their day acting as data-entry operators. This is slow, it creates errors, and it represents direct losses, because a delay in launching a campaign means missed impressions and client money spent inefficiently.

Why Conventional Automation Wasn't Enough

Media planning tools had existed in the market for years. But all of them only addressed part of the problem: they helped build a plan or generate a proposal, yet they never removed the core bottleneck, the manual negotiation between two parties. The moment actual deal terms needed to be agreed upon, the system stopped and handed control back to a human.

The issue wasn't a lack of automation per se. It was that existing solutions couldn't negotiate. They could prepare data but couldn't close a deal on their own. What was needed was an agent capable of acting autonomously: receiving a brief, forming a proposal, agreeing on terms, and activating a campaign without a human at every step.

This is where Fox placed its bet on AI agents, not merely as tools to speed up individual operations, but as full participants in transactions.

How the AI Agents for Fox AdStudio Were Designed

Fox transformed the Fox AdStudio platform by embedding a system of AI agents that support the complete campaign lifecycle. The key design decision: agents were built not as human assistants but as autonomous participants in deals.

Functionally, the agents cover several critical areas. Planning and buying: the agent automates audience discovery, reach forecasting, and placement recommendations, then independently manages the exchange of briefs and proposals between media planners and sellers. Agent-to-agent transactions: this is a distinct architectural layer. Fox built a trusted, governed environment in which an AI agent on the buyer's side and an AI agent on Fox's side can negotiate directly and close deals without a human involved in every approval round. Linear TV automation: this is where the biggest impact concentrates. Traditional linear television, with its slot negotiations, placement activations, and paperwork, is now processed by agents automatically.

Stefano Kim from Fox put the philosophy clearly: "Many are focused on automation, which is certainly important, but we felt we needed to solve for an authentic and secure method of agent-to-agent transactions." That distinction matters: the goal was not to make humans faster, but to remove them from the routine cycle entirely, leaving them only strategic decisions.

How the Rollout and Process Redesign Worked

The deployment was built around Fox AdStudio as the single platform into which agents were layered on top of existing workflows. The key decision was not to make teams learn a separate tool, but to integrate agents into the familiar interface where planners and sellers already worked.

In parallel, Fox had to address the trust question around autonomous transactions. Agent-to-agent deals are a new model for the market, and advertisers needed to be confident the system would behave predictably and securely. The architecture therefore included a governed environment with clear rules: what the agent decides on its own and when it escalates to a human. This removed the primary barrier, distrust of autonomous decisions on deals involving real budgets.

The rollout covers both streaming inventory and traditional linear TV. The latter is especially telling, because linear TV has historically been considered the least automatable segment due to the complexity of negotiations and the large number of manual steps involved.

Results

The numbers here are qualitative, but the economics behind them are real. A deal that used to take several working days now closes in seconds. This means Fox can sell the same volume of inventory at lower operational cost, and advertisers get campaigns live faster. Specialists freed from routine moved to audience analysis and strategic planning, work that actually drives campaign performance.

A separate result is scalability. Previously, growing deal volume required proportional growth in headcount. Now Fox can handle a larger number of transactions without expanding its team, because the bottleneck has been removed.

How to Apply This to Your Business

The media market is illustrative here, but the logic applies more broadly. If your business has processes where two sides spend days negotiating terms through emails, calls, and documents, that is a candidate for an AI agent. Where to start:

• Find your longest approval cycle. Any process where days pass between first contact and final decision because of manual information exchange is the first place for an agent. Count the hours, not just the steps.
• Separate routine from judgment. The agent handles data gathering, proposal generation, condition checking, and document flow. Humans make decisions in non-standard situations. A clear boundary makes the rollout predictable.
• Embed the agent in existing tools. The less the team needs to change its habits, the faster adoption happens. An agent layered on top of a familiar interface outperforms a separate system that requires onboarding.
• Settle the trust question upfront. Autonomous transactions trigger skepticism. Build escalation rules into the architecture from day one: what the agent decides independently and what it passes to a human. This removes the main adoption barrier.
• Measure scalability, not just speed. Fox's core argument was not just "faster" but "more deals without growing headcount." That is a different conversation with your CFO.

If this case sounds like what's happening in your company, our manager can help: he'll analyze your business and niche for free and point out where an AI agent would bring a real result in your case. Message the manager