At NAB, AWS says the AI conversation in broadcast has finally grown up

Amazon Web Services arrived at the 2026 NAB Show with an argument it has not always been able to make: that the artificial intelligence capabilities it is demonstrating on the show floor are already running in production at major broadcasters, not waiting to be tested or proof of concepts. 

The distinction matters at a show where AI has been a dominant theme for two years and where the gap between what vendors demonstrate and what customers actually deploy has been a persistent source of skepticism. This year, AWS came with named partnerships, shipping products and customer case studies that put specific workflows behind the broader agentic AI narrative.

Steph Lone, global leader for solutions architecture in media, entertainment, games and sports at AWS, said the shift toward agentic AI, systems in which multiple automated agents handle complex, multi-step workflows without manual coordination, reflects what broadcasters are actually asking for.

The conversation, she said, has moved from what AI could theoretically do to what it is doing right now in operations.

“When you think about why that is, customers are really trying to have more operational efficiency,” said Lone. “Having a host of agents being able to do a lot of these things makes a whole lot of sense to be able to get more product out the door.”

Fox and the vertical video problem

The headline partnership announcement at the show was Fox Corporation naming AWS as its preferred AI cloud provider, with the agreement covering Fox’s full portfolio including Fox Sports, Fox News Media, Tubi, Fox Entertainment and Fox Television Stations.

Fox Corporation names AWS as preferred AI cloud provider

The most concrete expression of that partnership addresses a workflow problem broadcasters have struggled with for years: how to produce vertical video for mobile and social platforms without building a separate production pipeline alongside the existing 16:9 broadcast operation.

AWS Elemental Inference, demonstrated at the show, is a fully managed service that applies AI in parallel with live video encoding, enabling broadcasters to automate 16:9 to vertical video creation in six to ten seconds. Fox Sports Digital and NBCUniversal are among the customers already using the service.

Advertisement

Lone said the urgency behind that capability is demographic.

“Vertical video is huge, and all of our broadcasters are really having to compete,” said Lone. “Consumer choice is endless, right? They’re having to compete with the likes of TikTok, Instagram, YouTube. This is the next wave of where consumers are today and our broadcasters need to be there at the moment.”

She cited data showing 88% of Gen Z consuming streaming content on mobile, placing vertical video squarely in the path of any broadcaster trying to reach younger audiences on platforms like TikTok, Instagram or YouTube. Broadcasters, she said, need tools that let them compete for that attention without doubling their production workload.

The Fox relationship predates the NAB announcement. Since 2022, the company has used the Amazon Media Replay Engine to automate recap videos for live events including the 2022 FIFA World Cup, and in 2023 upgraded its streaming infrastructure to support 4K with HDR and low latency while incorporating Amazon SageMaker, Amazon Personalize and Amazon Bedrock. The preferred AI cloud designation announced at NAB represents a formalization of that relationship rather than a new beginning.

Agents in the newsroom

Beyond sports production, Lone described a growing focus on newsroom workflows, an area where agentic AI takes a different form. Where sports production benefits from speed and automation in highlight generation and format conversion, newsrooms need help finding, assembling and publishing stories faster with fewer people.

AWS demonstrated a workflow at the show in which a group of agents searches for relevant content, assembles a rough cut and surfaces it for editorial review. The human editorial decision remains with the journalist. Lone said that human-in-the-loop element is deliberate, and that governance and guardrails around AI outputs are among the first things enterprise customers ask about.

Archive monetization is another area where Lone said customer demand is strong.

Broadcasters sitting on decades of content with inconsistent or missing metadata have historically needed significant manual effort to surface material for distribution or licensing. Lone said the ability to search a library using natural language, finding content by what it depicts rather than what it was labeled — is unlocking value that has been difficult to access.

Where it goes next

“We are the only ones that are investing in media services for the media company industry,” she said. “That was one of the reasons I came here, because we really pride ourselves on one industry expertise and the services that actually help our customers meet their moment.”

Lone described the direction customers are pulling toward as tools that remove operational friction between having content and getting it to an audience, across formats, platforms and markets simultaneously.

She pointed to adtech as one area where agentic AI is beginning to compress workflows that have historically taken weeks: ingesting an RFP, identifying available inventory, generating creative for smaller advertisers and moving toward execution with minimal manual intervention.

Advertisement

The customer demand driving that, she said, comes back to the same two words heard across the NAB Show floor all week.

“More efficiency, more scale, more security — all of those things,” she said.