The next stage of AI in broadcast planning is a team of specialised AI agents working together on a complete planning objective.
A planner can begin with an intention: “Prepare a film channel for an older audience. Use premieres in prime time, series in the afternoon, avoid frequent repeats, and identify any missing rights.”
From these instructions, AI agents analyse the available content, check licence availability, propose a channel structure, prepare the schedule, plan promotions, and identify decisions that require human input.
Through MCP (Model Context Protocol), the agents can work with the relevant data and functions available in PROVYS Sphere. The Model Context Protocol is used so that AI applications and agents can access external data, tools and workflows in a standardised way. An MCP server is then a specific server implementation that makes these resources and functions available to AI.
One objective, multiple roles
Complex planning work is divided among AI agents with clearly defined roles:
- A strategy agent interprets the programming objective.
- A catalogue agent identifies suitable content and programme versions.
- A rights agent checks whether the proposed use is permitted for the relevant channel, territory, platform, and period.
- A scheduling agent turns the approved strategy into a transmission plan.
- Other agents can focus on promotions, break patterns, EPG metadata, or financial implications.
The agents share the same operational context in PROVYS Sphere. They exchange results, identify conflicts, and return unresolved decisions to the planner.
For example, the strategy agent may request recent films for prime time. The rights agent may find that several titles are unavailable. The catalogue agent then proposes alternatives, and the scheduling agent updates the plan.
This is likely to be the practical model for AI planning: not full AI autonomy, but continuous cooperation between people and specialised agents.
Planning will become conversational
The interaction changes when planners no longer have to define every individual system action. Instead, they set the objective, priorities, and limits for the agents.
They can refine the plan in operational language:
“Reduce repeats in the evening schedule.”
“Use the available promotional inventory more evenly.”
“Show me which parts of the plan depend on rights that have not yet been secured.”
The agents use these instructions to work across content data, rights, schedules, programme versions, promotions, and metadata.
Proposed plans still require review, correction, and approval. A technically valid schedule may not reflect the intended channel identity, may overuse particular titles, or may apply a promotion strategy too aggressively. The result is not accepted blindly.
This points towards a realistic future for AI in broadcast planning. The planner does not disappear. Their role moves towards defining objectives, setting boundaries, reviewing proposals, resolving exceptions, and making editorial or commercial decisions.
AI handles more of the analysis and routine execution between these decisions.
Full autonomous broadcasting is not happening
The growing role of AI agents does not remove the need for a structured broadcast management system.
PROVYS Sphere continues to hold the operational data, validation logic, business rules, and planning workflows that the agents work with. The broadcaster remains in control of which data and functions are available to them and which actions require approval.
Agents can have flexibility when proposing programming strategies or comparing schedule variants. Rights validation, financial controls, approval steps, and schedule checks remain governed by defined system rules. People will continue to define the channel. AI will help turn that strategy into an operational plan.
Interested in the topic? Read also how structured data and workflows help overcome the challenges of AI in broadcasting.