How AI Is Personalizing the Live Sports Fan

I consume sports exactly the way that keeps sports executives up at night. During any given game, I'm cycling between the broadcast, a team subreddit, Twitter/X, a fantasy dashboard, and at least one group chat where someone is about to say something unhinged about a referee. I am not watching the game. I am orbiting the game through six simultaneous gravitational fields. And according to Scott Gutterman, chief digital officer at Next League, that chaotic multi-screen experience is actually the design brief for where AI in sports is headed.

Gutterman sat down with Front Office Sports for their Future of Sports series, and what comes through isn't a hype pitch — it's a pretty honest account of where the industry actually is versus where it wants to be, and the gap between those two things is instructive.

The "lean back" problem

Gutterman's core argument is that AI's most meaningful contribution to the fan experience is reducing what he calls the "lean-in" burden — the constant, active effort required to stay informed about what you actually care about in a sport. His illustration: a PGA Tour event with 100-plus players on the leaderboard, and you care about four of them. Right now, getting updates on your four requires you to be doing something — pulling out your phone, walking to the leaderboard, hunting through an app.

"What we really want," he told Front Office Sports, "is either people, through glasses like we were talking about earlier or maybe through an earpiece or even just picking up their phone and seeing on the screen through a live activity, here's how my five favorite players are performing right now."

That's a genuinely useful framing. The information architecture of most sports apps is still built around the assumption that you want comprehensive data, presented in a way that rewards study. What most fans actually want is their relevant slice of the action, surfaced automatically. AI changes what's technically possible there — models can learn your preferences fast enough to make that personalization feel ambient rather than configured.

The front office version of this is equally interesting. Gutterman describes AI letting teams move from reactive fan management to proactive engagement — imagine a fan who's attended multiple games already this season and is a registered app user; in Gutterman's framing, a system that knows that fan is likely to return might automatically surface a seat upgrade offer rather than waiting for the fan to request one. He's explicit that this isn't fully operationalized everywhere yet — organizations are still in the "preparing the data pipeline" phase — but he thinks AI will accelerate the pace considerably.

This is also where the agentic AI deployment story gets interesting: autonomous agents that don't just surface recommendations but act on them — booking the upgrade, applying the discount, initiating the outreach — are moving from experimental to operational across parts of the sports business right now. Gutterman's vision of proactive fan management is essentially describing agent behavior; whether sports organizations are ready for that handoff of decision-making is a separate and genuinely open question.

The privacy thing, and why I'm not fully convinced by the industry answer

When host Baker Machado asked about the privacy tension — how do you make fans feel like they're getting value from their data rather than just being surveilled — Gutterman gave the answer I expected: make the experience good enough that consent feels obvious, get your data infrastructure in order, be transparent about what you're doing with IP and content rights.

That's the right framework in theory. It's also the framework every platform deploys before the thing goes sideways.

Here's my actual concern, and it's not unique to sports: the personalization-to-surveillance gradient is slippery in ways that don't require bad intentions. A sports app that knows your location, your purchase history, your favorite players, your attendance patterns, and your in-app behavior is holding a remarkably complete behavioral profile. That's useful for sending you a seat upgrade offer. It's also useful for dynamic pricing that extracts maximum willingness-to-pay from exactly you, specifically, in real time. Those are not the same thing, and they're enabled by identical data infrastructure.

Gutterman doesn't address that second use case, which is fair — it wasn't what he was asked — but it's worth naming. The sports industry's privacy argument right now rests almost entirely on "we'll use this to give you better experiences," and I think fans who've watched streaming services use the same pitch before quietly optimizing for churn prevention and upsell should ask a few follow-up questions. What data, retained how long, shared with whom, for which purposes? "Earn consent through value" is a good starting position. It's not a complete answer.

That said: I don't think the answer is to refuse the technology. The lean-back experience Gutterman describes is genuinely better than what currently exists. I just think the consent architecture needs to be built before the personalization layer, not retrofitted after it's already running.

Spatial computing: the Ray-Bans are doing more work than the Vision Pro

The spatial computing section of this conversation is where Gutterman is most candid. He's bullish on lightweight glasses (your Ray-Ban Meta collab, your inevitable competitors) and honest that the full headset experience is still a proof-of-concept game — genuinely impressive, not yet something fans are choosing at scale. He said the PGA Tour worked on spatial computing experiences dating back to some of its earliest immersive streaming experiments, eventually building up to its Apple Vision Pro work, and a lot of that effort was "proof of concept and figuring out how this environment works."

His timeline for something solid: three to five years, specifically for the glasses form factor, and with a hard requirement that it be social. This is the part I find most convincing. The reason VR sports experiments have repeatedly underperformed isn't technical — it's that watching a game alone in a headset is actively worse than the group text experience I described at the top. Sports fandom is about shared witnessing. You want to turn to someone and say "did you SEE that?" The glasses form factor at least preserves the ability to do that without looking like you're prepping for surgery.

Gutterman's hypothetical: you're at a Yankees game, someone hits a 400-foot home run, and the distance appears on your glasses before you've even thought to ask. That's a different relationship with in-stadium information — additive to the live experience rather than distracting from it. Whether the form factor gets there in three to five years or closer to ten is genuinely unclear, but the direction feels right.

The distribution question nobody has fully solved

One of the more honest moments in the conversation is Gutterman acknowledging that sports organizations spent two decades trying to drive everyone to owned apps and websites — and now they're accepting that the fan is wherever the fan is, including on ChatGPT checking last night's scores or TikTok learning about a driver's personality they'd never encounter in the official NASCAR app.

"If you want to get to know Joey Logano," he said, "you're probably going to get to know his personality better on TikTok than you are necessarily on the app or website where you're going to get great information but not necessarily the personality."

That's a real tension between reach and control. The owned platform is where the authenticated data lives, where the personalization runs, where the revenue gets captured. But the discovery moment — the thing that makes someone a fan in the first place — increasingly happens on platforms the league doesn't own and can't fully instrument. Gutterman's answer is essentially: accept it, optimize your content IP for distribution everywhere, and use the owned platform for what it uniquely offers (speed, accuracy, depth). That's probably right as a strategy. It also means league platforms are increasingly competing for attention with the same algorithmic feeds they're trying to use as a top-of-funnel.

Where Gutterman is right, and where I'd push back

He's right that AI as a background layer — ambient, proactive, reducing friction — is more interesting than AI as a feature you have to engage with. The "AI tab" in a sports app is the wrong deployment. AI that already knows which hole Rory McIlroy is about to tee off on and routes you there is the right one.

He's right that infrastructure flexibility ("infrastructure as code," as he puts it) matters more than any specific technology bet right now. The teams over-indexing on a single hardware platform will be rebuilding in five years; the ones building adaptable pipelines will just swap the layer.

Where I'd push: the "everyone can be a VIP" framing is appealing and probably partially true, but it assumes organizations will use personalization data to give fans more value rather than to capture more of their surplus. Those are different business decisions, and the technology doesn't make them for you. AI that knows you're a high-frequency attender could give you a free ticket — or it could give you a price that extracts every dollar your revealed behavior suggests you'd pay. Gutterman works with leagues on this, so he's oriented toward the former. The actual outcome will depend on what the business incentive structure rewards.

That question — who does the personalization ultimately serve — is the one I'll be watching as this space matures. The technology to make every fan feel like a VIP is being built. Whether that's what it gets used for is still being decided.

Yuki Okonkwo covers AI and machine learning for Buzzrag.