We speak with Steph Lone, Global Leader, Solutions Architecture, Media & Entertainment, Games and Sports at AWS, about the broader industry trends reflected in partnerships such as FOX Corporation’s decision to name AWS its preferred AI cloud provider

*This content is included in the June issue of TM BROADCAST. You can access the full magazine here.

The recent announcement of FOX naming AWS as its preferred AI cloud provider may appear, at first glance, to be just another technology partnership. Yet, according to AWS, the agreement reflects a broader transformation taking place across the media industry.

“Our partnership with FOX Corporation spans Fox News Media, Fox Sports, the Fox broadcast network, and the FOX One direct-to-consumer platform, which serves 300 million monthly active users,” explains Steph Lone, Global Leader, Solutions Architecture, Media & Entertainment, Games and Sports at AWS, to TM BROADCAST. “It’s indicative of a larger shift we’re seeing amongst media companies today.”

That shift goes beyond infrastructure. Rather than engaging with technology vendors on a project-by-project basis, media organizations are increasingly seeking long-term partners capable of supporting transformation across the entire content lifecycle, from creation through distribution and monetization.

“As FOX’s preferred AI cloud provider, AWS provides the speed, scale, and agility they need to modernize faster across their entire business portfolio,” says Steph Lone. “The relationship extends far beyond a typical transactional vendor to a more strategic, enterprise-wide technology alliance. It signals a shift in partnership culture that the industry will start to see more broadly across the board.”

Rather than engaging with technology vendors on a project-by-project basis, media organizations are increasingly seeking long-term partners

From experimentation to operational impact

If partnerships are evolving, so too is the role of artificial intelligence within media organizations. While generative AI often dominates headlines, AWS points to a growing number of use cases where AI is already delivering measurable operational benefits. These are typically workflows characterised by high-volume tasks, structured inputs and clearly defined outputs.

“Media companies are now seeing real operational impact with AI in areas that have high-volume tasks with clean inputs and clear outputs. These might include highlight and vertical video generation, video intelligence, content production, and system troubleshooting, among others.”

FOX Sports provides one example. Through AWS Elemental Inference, the broadcaster can automatically transform traditional 16:9 live sports feeds into vertical formats for social media consumption within 6-10 seconds, enabling near real-time distribution across digital platforms. “Since video is processed once but optimized for multiple functions, it reduces costs by 34%. That makes a huge difference, considering nearly 90% of Fox Sports digital content is consumed vertically.”

 “Personalization in live sports is moving from post-event recommendations to in-game, multi-format, multi-platform delivery”

Other organizations are applying AI to entirely different challenges. Bloomberg Media, for instance, has built an AI-powered video intelligence platform on AWS capable of extracting value from its 13-petabyte archive. According to Lone, the system combines vision language models, multimodal embeddings, and knowledge graphs powered by Amazon Bedrock. “AI agents extract metadata, correlate assets via federated search, and assemble platform-optimized clips for distribution across 48 streaming platforms, reaching 60 million monthly viewers. This dramatically reduces time-to-market and unlocks new revenue from previously unstructured content.”

Sports organizations are also finding practical applications. “The PGA TOUR features timely coverage on every player using a multi-agent AI system based on AWS that produces over 1,200 articles per week at a 95% cost reduction,” highlights Lone.

Formula 1, meanwhile, has deployed a multi-agent root cause analysis system on Amazon Bedrock to support the more than 500 applications behind F1 TV Premium and race-day operations. “Supervisor and specialist AI agents analyze CloudWatch logs, correlate incidents, and auto-generate tickets, allowing F1 to cut initial triage from 45 minutes to 4 minutes (a 95% reduction) while predicting and preventing issues before they reach fans.”

Taken together, these examples suggest that the industry is moving beyond experimentation and into a phase where AI adoption is increasingly tied to measurable business outcomes. “While some companies are experimenting with creative applications, human review remains essential,” adds Lone.

“Since 88% of Gen Z consumes streaming content on smartphones, vertical video is also a key consideration”

Personalization moves into the live environment

Another area undergoing rapid transformation is personalization. Historically, personalization largely revolved around content recommendations after an event had ended. Today, according to AWS, the objective is becoming much more ambitious. “Personalization in live sports is moving from post-event recommendations to in-game, multi-format, multi-platform delivery,” says Lone.

The implications extend beyond recommendations. “The UI, the overlays, the ads, and the companion content all adapt dynamically based on the viewer, their device and location, and what is happening in the game.”

This evolution reflects changing audience consumption habits, particularly among younger viewers who increasingly engage with content through mobile devices and social platforms. “Since 88% of Gen Z consumes streaming content on smartphones, vertical video is also a key consideration.”

Examples of this shift are already emerging across the industry. According to AWS, media companies are increasingly combining real-time content formatting with AI-powered content discovery to support more personalized experiences at scale.

“Media companies are now seeing real operational impact with AI in areas that have high-volume tasks with clean inputs and clear outputs”

The Bundesliga’s official application, for example, uses Amazon Personalize to deliver individualized content experiences based on user behaviour and preferences. “Using Amazon Personalize, the Bundesliga (DFL) Official App generates individualized content for millions of active users each season based on their favorite teams, players, and behavior. It’s helped increase article reads by 67% per user and improved the time users spend in the app by 17%.”

Amazon Prime Video is pursuing a similar objective through generative AI. “Amazon Prime Video has replaced traditional genre-based browsing with personalized content recommendations using generative AI powered by Amazon Bedrock.”

Meanwhile, broadcasters are also exploring ways to personalize content while events are still taking place. “With AWS Elemental Inference, broadcasters like Fox Sports can publish social-optimized vertical clips while live events are still in progress.”

“Since video is processed once but optimized for multiple functions, it reduces costs by 34%”

Delivering this type of real-time personalization, however, requires infrastructure capable of meeting the strict demands of live production. Scalability, latency and reliability remain critical considerations, particularly as AI becomes embedded in operational workflows. “Scalability, latency, and reliability are non-negotiables for any broadcaster, and they become even more critical when you introduce AI into live workflows.”

According to AWS, services such as AWS Elemental Inference are designed to run AI processing in parallel with video encoding, keeping latency within a 6-10 second window while applying multiple AI functions simultaneously, including vertical reformatting, highlight detection and metadata enrichment.

“The more features you enable with AWS Elemental Inference, the more cost-efficient it becomes. Since it’s a fully managed, autonomous service that integrates with existing AWS Elemental MediaLive workflows, broadcasters do not need to provision, configure, or maintain infrastructure or modify their current video architecture to adopt AI capabilities. In terms of reliability, AWS continues to have the strongest track record for availability, reliability, and security amongst cloud providers.”

More broadly, the company argues that AI should enhance live production workflows without adding complexity or operational risk. “As AI adoption accelerates, the challenge is no longer whether these capabilities are technically feasible, but how to deploy them at scale while maintaining the reliability standards expected in broadcast environments.”

“As AI adoption accelerates, the challenge is no longer whether these capabilities are technically feasible, but how to deploy them at scale while maintaining the reliability standards expected in broadcast environments”