🧠 How Broadcasters Can Implement AI Without Losing Viewers' Trust

Despite regulatory initiatives—such as the OECD’s updated principles or the European Union’s AI Act (EU AI Act)—most technologies used in broadcasting and live transmissions fall into limited or minimal risk categories. Nevertheless, customizable flexibility and algorithmic transparency are becoming essential conditions for maintaining audience trust.

Karoline Cardoso, Vice President of Product Management at Vizrt, points out that there is no single scenario for using AI in content production. The top priority remains human oversight at every stage of the process.

The degree of algorithmic involvement can vary depending on a newsroom’s specific needs:

• Full Automation: AI agents trained on editorial guidelines independently select media assets, generate text based on a journalist’s prompt, and push the story live to air.
• Hybrid Control: Algorithms prepare a news draft or compile illustrative media, but final approval and editing remain strictly with a human editor.

As systems become more complex, the role of self-verification and audit logging grows. When handling complex or potentially risky tasks, AI agents must be capable of “explaining” the rationale behind their choices—for instance, justifying why a specific illustration fits a given text.

To ensure safe deployment, developers are implementing centralized control centers. In these systems, users set strict parameters in advance. If an algorithm’s next step crosses those pre-defined boundaries, the system automatically halts the process and requests human operator confirmation.

Karoline Cardoso emphasizes that integrating AI does not require secrecy. Adhering to safety protocols while remaining open with the audience allows broadcasters to boost production efficiency without compromising viewer loyalty.