AI In Broadcasting: Practical Fixes Over Hype

A recent webinar on AI and media workflows brought together executives from Graham Media Group, Morgan Murphy Media, Fox Television Stations and Ross Video to cut through the hype and focus on what is changing in day‑to‑day operations. The consensus: use automation to handle routine work, ensure tools interoperate cleanly, and keep final editorial judgement firmly with humans.

Moving beyond isolated pilots into production environments is now the aim at Graham Media Group, according to Director of Transformation Michael Newman. Progress is uneven across teams, he noted, but the direction of travel is to embed AI inside existing systems rather than treat it as a side project. That shift puts fresh emphasis on access controls, reliability, metadata discipline and support processes for broadcast engineers.

Morgan Murphy Media's Chief Innovation Officer, Colin Benedict, set out the newsroom case simply: let machines take care of SEO, tagging, clipping and repackaging so journalists can do more journalism.

"Do we really want our journalists to be focused on tools? Or do we want them to be focused on finding compelling stories, talking to people in the community, being creative in their storytelling?" Colin Benedict – Chief Innovation Officer – Morgan Murphy Media

Fox Television Stations' Tim Joyce underlined that adoption depends on culture as much as technology. In live broadcasting, "mostly right" can still cause real issues, so the smart place to begin is with low‑risk, high‑value applications that measurably improve efficiency without touching critical playout. Examples discussed included captioning, asset tagging, facial tracking, semantic search and script‑aware graphics suggestions—jobs that are tightly connected to the workflow and easy to evaluate.

From the vendor side, Ross Video's Jenn Jarvis stressed that AI should only be used where it is the best fit for a concrete production problem.

"We're definitely not using AI just for the sake of it. But when we're trying to solve a problem and AI is the best tool for the job, it makes sense.
Jenn Jarvis – Manager, Solutions – Production Workflow at Ross Video

Integration repeatedly surfaced as the real bottleneck. AI agents sound impressive, but deliver little unless surrounding systems can exchange data and context. Siloed tools and patchy metadata undermine results. Newman summed up the operational need: "To have a durable, consistent metadata across all of these different platforms is really important to be able to take the most advantage of what is available for a specific story." Michael Newman – Director of Transformation – Graham Media Group

Breaking news provides the clearest stress test. A single development can rapidly spawn live streams, social posts, web updates, graphics, alerts and archive pulls. The panel's point was not merely that AI could tidy up after the fact, but that better orchestration and shared story data from the outset make those assets easier to track, repurpose and publish across platforms at speed.

Trust and governance were non‑negotiable. No one argued for removing human judgement; quite the opposite. If automation is going to touch editorial or production workflows, teams need traceability: what content was used, what the AI changed, what a person approved, and why a decision was taken. Well‑placed human checkpoints—particularly where on‑air output, public publishing or intellectual property are involved—are the practical basis of trust, not red tape.

Panelists also discussed AI's role in raising editorial quality. Fox is testing tools to flag questions around fairness, balance and language, while Morgan Murphy is exploring prompts that surface the kinds of follow‑ups an editor—or a viewer—might expect. With audience metrics piling up faster than teams can interpret them, analytics that distil meaningful insight were another recurring theme.

There was clear guidance on where to tread carefully. Digital producer roles will evolve as automation takes on more packaging and distribution work. By contrast, master control was cited as a poor place to trial autonomy. Start where tasks are repetitive, the risk is low and the benefits are obvious, and expand from there. As Benedict put it: "We look at it through the lens of, 'How can we use those roles to create content that's more meaningful to our audiences?'
Colin Benedict – Chief Innovation Officer – Morgan Murphy Media

For engineers, the bottom line is straightforward. Success will not hinge on clever prompting; it will come from robust plumbing—clean, consistent metadata; open, well‑documented interfaces; useful logs; sensible permissions; and systems that can be trusted when the red light is on. With that foundation, teams can apply automation confidently to real newsroom workflows rather than keep it in the experimental bucket.

www.rossvideo.com/