Content creation and rights, YouTube archive, production workflows, investment, and governance were primary topicsDevelopments in the world of AI and generative AI are happening at a profoundly fast clip, and the question facing nearly every sports organization on the planet is how not only to keep up with the pace of change but also to ensure a corporate strategy designed to protect intellectual property, employees, and even fans. According to a panel discussion at the recent Sportel conference in Monaco, the industry has moved past the testing phase.
It's an interesting space where we are now moving from experimentation to actual productization, which is content production, operations, and rights, explained Prashant Khanna, head of rights, innovation, and Cricinfo business, JioStar, who led the discussion. We have moved from what was good to have' to making sure that it is part of our fabric in some way.
Khanna described a progression through three stages: traditional AI, generative AI, and agentic AI. Traditional AI, he said, provides the baseline. Generative AI makes derivative content more compelling while leaving the core live broadcast intact. Agentic AI helps decide what each fan should see.
Generative allows us to look at how we can make that derivative content very exciting, Khanna said, which is the same experience in terms of live sport but making sure it is being delivered to fans with a lot of personalization.
He also noted that fan behavior is changing the equation. Fans are increasingly engaging with their favorite sports much differently than just sitting down and watching a full match or game. And that, in turn, increases the need to use AI and generative-AI tools to help serve fans' widening needs.
It's no longer just a 90-minute or 120-minute live game, Khanna pointed out. Some fans are experiencing it first on social. - their first entry point - and some are watching it through a 10-second reel, a 30-second reel, a 10-minute highlight, and then moving into the live game.
Squeezing More Value From YouTube Archives
And then there is the world of YouTube, which is increasingly home to sports content. BuzzMyVideos helps rightsholders maximize their content globally and builds technology platforms that scale the work of improving revenue and reach on YouTube.
BuzzMyVideos CEO Paola Marinone, a YouTube alum, compared the platform to a vast reference library. YouTube is basically the Wikipedia of content as we speak. You need to make sure you maximize it at best, and you need technology to do that.
Her focus is on material that rightsholders have already uploaded, and she said the platform connects through YouTube's API so that thousands of videos can be optimized with a few clicks.
Much of the value, she noted, sits in the back catalog. We were running some analysis the other day, and it turns out that 50% of sports content watched on YouTube is archive material. A game from 10 years ago actually has value on YouTube because it keeps on giving.
At Sportel Monaco, a panel comprising (from left) JioStar's Prashant Khanna, BuzzMyVideos' Paola Marinone, Ekstraklasa Live Park's Marcin Serafin, SURJ Sports Investment's Danny Townsend, and Sports Video Group's Ken Kerschbaumer discussed how each organization is approaching AI.Asked what optimization means in practice, Marinone said it covers everything surrounding the video, not the footage itself, and the goal is to influence YouTube's recommendation algorithm. It's the title, the thumbnail, the description - everything that is not the video itself. You might have brilliant content, but, if [fans] don't find it, it's not going to work out.
Ekstraklasa Live Park COO Marcin Serafin offered a different view of archives from the league's perspective. Live Park, a TV-production company owned by Polish football league Ekstraklasa since 2011, produces the TV feed for every top-flight league match and runs the league's official YouTube channel and archive. Highlights, he said, generate far more traffic than older games.
The archive is nice to have, but it's not even a core product for us, honestly, Serafin said, adding that the league tried building its own archive infrastructure but chose not to invest heavily in something it cannot sell.
Remote Production, Cost Control, and the VAR Question
The panel next looked at how AI is being deployed in production workflows. Live Park's priorities, Serafin said, are cost-effectiveness and solving a staffing problem.
A lot of companies are missing people: people are getting older, and the younger generation feels they have better things to do on weekends than travel to games all around Poland, he explained. That is why the company is pursuing remote production from a single headquarters, along with equipment that includes what he called AI-ready features.
On the generative side, Serafin said, the company keeps spending modest. Using Adobe features, ChatGPT and ElevenLabs, we spend no more than 200 or 300 per month, and we get better content, mostly for YouTube.
He also raised a thorny question about a new feature in the league's EVS replay system. The extra motion capability creates super slow-motion by interpolating additional frames, which is cheaper than deploying extra super-slo-mo cameras. But the added frames are generated, not captured.
We are waiting for official approval because of European law, Serafin said. Nobody knows if it's real AI or not, because the feature interpolates and adds extra frames.
The concern is video-assistant-referee (VAR) reviews. Live Park has not yet put the feature to use while it awaits an official statement from EVS. Nobody knows what's going to happen if it's a handball or not in the penalty area in the last minute, S
