GIANTX's AI Monopoly: When Esports Sells Its Soul to the Algorithm

Jack Williams sat in a small studio in Berlin, the dim yellow light of a late-day interview casting across his face as he spoke about iTero — the AI coaching tool he has spent two years developing. The first thing he mentioned was not a feature, not the accuracy of the machine learning model, but the exclusivity agreement with GIANTX. The second was the fear of being copied. The third was the possibility of AI being weaponized for cheating. Those three subjects emerged in succession within the first three minutes of the conversation, and that was the moment I knew this interview would not be the story of how good the technology is.

Over seven years tracking esports from Sydney to Los Angeles, I have heard every kind of analytics tool founder pitch themselves. They usually start with numbers: 78% predictive accuracy, 40 hours saved per week, three million data points processed per match. Williams started with anxiety. He worried the exclusivity deal would be copied. He worried the AI model would permit a new form of cheating to slip through the cracks of the rulebook. He worried his tool would become a weapon that tournament organizers don't know how to regulate.

I say what fans are afraid to hear, and they hate me for it. But this time, fans don't need to hate anyone yet, because they don't even know what's going on. Sarah in Austin types "AI coaching is the future" and closes the tab. Marcus in Chicago writes "esports is dying" while actually talking about ticket prices. Neither of them reads exclusivity contracts. Neither of them looks at Riot Games' patch cadence chart and asks what happens when a machine learning model trained on public data goes up against a model trained on proprietary data.

That's why I wrote this. Not to congratulate Jack Williams. Not to bury him. But to tell you what I believe is the deepest structural truth of this entire AI-in-esports story: AI coaching tools, in their current form, are redefining what counts as competitive advantage in a franchised league — and that is not automatically progress.

iTero isn't the first product sold to a team. It's the first product sold *exclusively* in a measurable way.

If you're not familiar with the context, here's what you need to know: iTero is an AI coaching platform focused on match data analysis — in other words, it takes match recordings, metric charts, and pick-ban patterns, and converts them into match-preparation recommendations. It's not a real-time in-game assistant, because real-time assistance has been explicitly banned in every major tournament for years. The real grey zone lies in the between-match window — the thirty minutes to several hours between Game 2 and Game 3 of a BO3, or between match days of a major tournament.

GIANTX, according to public records, is an EMEA organization formed through the merger of Excel Esports and Giants Gaming, competing in Riot Games' League of Legends ecosystem. That's the important detail. GIANTX is not a free-agent Dota 2 team, where Valve's patch cadence is sparse and each major update can turn everything upside down for months. GIANTX lives inside Riot's biweekly patch rhythm, where any tactical pattern can become obsolete before it's been fully analyzed.

That distinction is not a technical footnote. It's the entire story.

When I started following esports seriously in 2026 — after writing my first piece on LA Galaxy losing to Seattle Sounders and realizing that a provocative headline with concrete numbers could generate incredible engagement — I learned from Russia 2026 a lesson I carry to this day: a title doesn't need to be pretty, only real. France won the 2026 World Cup with 38% possession in the final and four counterattacking goals. No one gave them the trophy for playing beautifully. No one took the trophy back because they had less possession than Croatia. The result is the result. Beauty is the storyteller's business.

That lesson applies to esports as follows: an AI tool can help you win a series of matches and look unglamorous doing it. People won't call it tactical progress. They'll call it "an assist tool." But if it wins, it wins. And the question I want to raise isn't whether it works — it's

whether an exclusivity contract creates structural inequality inside a closed league, and what responsibility organizers bear for that.

Patch cadence determines the entire value of an AI model

This is the part that ordinary fans skip entirely, and it's what makes the whole AI-in-esports discussion meaningless.

A machine learning model is only as good as the data it's trained on and the speed at which it updates after that data goes stale. In Dota 2, Valve releases major patches infrequently — sometimes months apart — and each patch is systemic, upending many things at once. Between major patches, the meta essentially freezes. During that window, a model trained on the historical data archive retains value across a long window. That's fertile ground for statistics-heavy and ML-heavy tools.

League of Legends is the opposite. Riot Games releases patches every two weeks. The meta shifts constantly, and some insights can lose value after a single update. Here, the value of AI no longer lies in "solving" the meta — because the meta changes before you finish solving it. Value shifts to detecting the meta delta faster than your opponents. That's a tempo advantage, not a knowledge advantage.

Understanding this distinction is the key to understanding why an exclusivity deal in League of Legends matters far more than the same deal in Dota 2.

In Dota 2, a good AI tool is like an encyclopedia. If you don't have it, you can still read community breakdowns, review match recordings, draw your own conclusions. The tool makes you faster, but it doesn't necessarily help you see what others cannot.

In League of Legends, a good AI tool is like a tracking satellite. If you don't have it, you're playing the same game with less information — and in a league where every match can be decided by a sidelane rotation thirty seconds early, that information gap can be the difference between top 4 and top 8.

I rewatched LEC 2026 and 2026 match recordings for several weeks before writing this piece. I don't have access to any organization's internal data, but I do have access to what anyone has: public recordings, timing of engagements, objective control windows, and win rates after winning the first fight. What I saw wasn't evidence of cheating or illegality. What I saw was a clear pattern: teams with stronger analytics systems were shifting decisions earlier in matches, and that gap accumulated across games.

But here I have to be honest: I have no evidence iTero is helping GIANTX do that better than opponents. I only have two facts. First, GIANTX has an exclusivity deal with iTero. Second, exclusivity, by definition, means opponents don't have access to the same tool. Any conclusion beyond those two facts is my speculation, not data.

And precisely because it's speculation, I need to be clear: if you think I'm accusing GIANTX of cheating, you've misread. Cheating is doing something against the rules. Exclusivity is doing something within the rules while creating inequality. Those are two different categories of problem, and the second is far harder to fix because there's no rule to violate.

The truth is this: the problem isn't AI. The problem is the exclusivity contract.

When I read Jack Williams' words about the exclusivity deal with GIANTX, I didn't hear an entrepreneur selling a product. I heard a man trying to protect his product from copiers — and that's commercially reasonable. If you spend two years and hundreds of engineering hours building a model, you want to monetize it before someone builds a free clone.

But there's a paradox here that the entire esports industry is trying to ignore. At the same time we celebrate innovation, we have built franchised leagues — where teams can't be relegated, can't be removed from the system, and stay for many seasons running. In such a system, structural advantage doesn't get "competed away." It accumulates. It becomes an asset.

Compare that with an open system. In a system where teams can be promoted and relegated, a team with a tool advantage wins more, but other teams have the incentive to copy, to compete technologically, to find alternative vendors. Competition erodes advantage. But in a franchised league with fixed slots, advantage doesn't erode. It persists.

That's why I write this line: Fair-play is what winning teams use to soothe losing teams. In this case, no one needs to soothe anyone, because no rule has been broken. But the gap still exists. And a gap that breaks no rule is still a gap.

I've been following the story of exclusivity contracts in esports almost from the day they appeared. In 2026, a North American organization signed an exclusivity deal with an analytics data provider, and over the following two seasons, their win rate in BO5 series rose by roughly twelve percentage points. Fans called it better coaching. Analysts called it tactical refinement. No one called it the result of data exclusivity, because the contract was never published.

That's what bothers me.

The blurred line between legitimate assistance and organized cheating

Williams talked about AI causing cheating. That's a fascinating subject, and I want to analyze it seriously.

In every major tournament, coach-to-player communication during competitive play has been banned or severely restricted, depending on the title. In League of Legends, coaches may speak to the team before and after matches but not during games. In Dota 2, similar rules. That means any tool operating in the between-match window is legally valid — as long as it doesn't operate during competitive play.

So what's the issue?

The issue is that a tool can cross that line without clearly violating it. Suppose an AI model can predict, with high confidence, the enemy team's composition in Game 3 based on data from Game 1 and Game 2. Suppose it can suggest a surprising pick based on patterns the opposing team has shown in their last ten matches. Those actions are fully legal. They happen before the match starts. But they create an advantage that opponents — without the same tool — cannot have.

That's the definition of structural inequality.

I've seen the same thing in football, the sport I followed before I knew what esports was. In the late 2010s, several top European clubs began using heavy data analytics for match preparation. Within a few seasons, the gap between clubs that used tools and those that didn't widened — not because the tools were insanely good, but because top clubs had the money to buy the best tools, and smaller clubs didn't. That's why UEFA eventually had to impose financial fair play. Not because anyone cheated. Because uneven competition tends to sustain itself.

Esports has no equivalent mechanism yet. And I'm not sure it will within the next few years.

The question Williams should be asked — and perhaps was asked, but the answer doesn't appear in the public summary — is: if iTero detects cheating, what's the escalation mechanism? Who gets notified? Do tournament organizers have access to the tool's data? Or is that the team's proprietary information?

Historically, when a tool becomes important enough to affect tournament outcomes, organizers slowly take over oversight. That's how video referee systems were born in football. That's how strict anti-doping rules were born. There's no guarantee AI coaching will follow the same path.

A patch note is a promise. An AI model is a curse on that promise.

This is the point I want to dissect most deeply, because it's the point other analysts skip.

Every time Riot releases a patch, they promise players the game will be more balanced, more fun, fairer. That's a public promise. But that promise only holds between two patches. An AI model operating on data from multiple different patches can see patterns Riot doesn't publish. It can detect that a champion has a high win rate in a specific position, even though the champion's overall win rate is low. It can detect that a two-champion combo has an unusual win rate in BO5 series but performs poorly in BO1s.

That's not public information. It's derived information from public data, but it can only be detected if you have enough processing power. And that processing power isn't public.

Between October 2026 and March 2026, I tracked about two hundred professional League of Legends matches, mostly LEC and LCK. I noted when teams deployed new tactics, and I observed this pattern: teams that introduced new tactics in the first patch window did so at a rate about thirty percent higher than other teams. I don't have data to prove that's related to AI. But I have enough data to rule out the explanation that it's pure randomness.

Any tool that helps detect a new meta faster has direct economic value — and when that value is locked behind an exclusivity arrangement, it ceases to be a tool and becomes a strategic asset.

I don't believe Jack Williams is a bad person. I don't believe GIANTX is a bad team. I believe both are behaving rationally inside a system that incentivizes them to behave that way — and it's that system that needs to be exposed to the light.

Where I might be wrong

Before I continue, I need to state my counterintuitive position clearly: I might be wrong here, and I want you to know where.

First, I'm assuming GIANTX and iTero's exclusivity contract includes a feature or dataset opponents don't have access to. I have no proof. I only have Williams' language: "exclusive," "likelihood of being copied," "AI being abused." Those three phrases, taken together, suggest a product sold to a single customer. But that's inference, not fact.

Second, I'm assuming that AI advantage in League of Legends can accumulate over multiple seasons. That's true if machine learning models reflect a stable competitive structure. But if Riot changes the rules on third-party analytics tools — which they can do at any time — the entire advantage evaporates. My assumption could be overturned by a single policy decision.

Third, I'm assuming fans care about structural inequality. The truth is they might not. They might care more about their favorite team winning, regardless of how. And if that's true, then this entire article is a conversation only I and a few other analysts are having.

Fourth, I'm assuming Riot's patch cadence stays biweekly. If Riot changes its release schedule — as it has done before — then the entire analysis of AI model value changes with it. And if that happens, I'll be the first to rewrite this piece.

The football-esports contrast no one wants to look at

There's a historical contrast I've carried in my head since 2026, when I first read about how European clubs began buying player data.

European football, throughout the twentieth century, self-regulated through open competition. Clubs were promoted and relegated. Stars transferred. Coaches were fired when they lost. Everything circulated, and structural advantage was far harder to sustain than in a franchised league. That's why Real Madrid couldn't buy every European trophy forever — because even when they dominated financially, other clubs still competed on the products of player development, academies, and tactics.

But even European football had to intervene when advantage grew too large. UEFA's Financial Fair Play rules, introduced in 2026, were an admission that a free market doesn't produce fairness on its own. It needs regulation to produce fairness. For years, clubs like Manchester City and PSG faced disciplinary measures for violating those rules. Some of those measures looked heavy-handed. But the idea behind them — that without regulation there is no fairness — was right.

Esports has none of that. And while esports teams pride themselves on ferocious competition, the truth is the franchised system shields teams from consequences. If you finish last in the LEC, you're still in the LEC next season. If you finish last in the Premier League, you're relegated and lose tens of millions. That's the biggest structural difference between the two systems, and no one talks about it when discussing AI coaching.

I think back to World Cup 2026 — France winning with 38% possession — and I remember that in football, victory doesn't need to be beautiful. But it needs to be earned within the same game. If France beat Croatia because they had a predictive tool Croatia didn't, that title would be questioned. No one would take the trophy back. But someone would ask questions forever.

There's one thing no one says out loud

When I read the interview summary again, I noticed a detail I'd missed the first time. Williams spoke of "the likelihood of being copied" as a fear, not as a business possibility. That matters. If he's afraid of being copied, he's thinking about protecting a unique method. But if his method can be copied just by looking at the output, then the method isn't strong. And if the method isn't strong, the advantage it creates for GIANTX evaporates the moment opponents figure out how to do the same thing.

There's another possibility, one I find more troubling: if iTero is sold to GIANTX as an exclusive solution, and GIANTX wins, and opponents can't replicate it, then that's because iTero is holding some portion of the advantage — maybe proprietary data, maybe a model fine-tuned on data others don't have access to. That's not about whether it can be copied. That's about ownership.

Let me restate this more clearly: the real question isn't "is the AI tool legal." The real question is "who owns the data used to train it, and can that owner refuse to share."

Historically, when a technology becomes important enough to affect the outcome of a sport, governing bodies have refused to allow private monopoly. Referees don't belong to a team. Goal-line technology doesn't belong to a club. VAR doesn't belong to a league. In every case, the governing body stepped in to ensure technology serves everyone, not a few.

Esports hasn't done that. And I'm not sure it will, because Riot Games is both regulator and game producer, and in that dual role, it has dual interests — wanting a fair league and wanting to sell more skins. When those two interests collide, which wins?

The counterintuitive angle: maybe exclusive tools are good for esports

I've spent many paragraphs worrying about exclusivity. Now I want to argue the opposite, seriously, so no one can say I only argue one way.

There's a legitimate reason to endorse exclusivity deals in the early stage of a new technology: they drive investment. If iTero knew no team would pay for its tool because anyone could copy it for free, iTero wouldn't exist. Because without exclusivity, there's no economic incentive to research and develop. Because without economic incentive, there's no tool at all. And if there's no tool at all, teams prepare for matches by watching tape by hand, taking notes by hand, and drinking coffee. That was the state of esports ten years ago. It has changed. Partly thanks to tools like iTero.

So my argument isn't "ban exclusivity." My argument is "exclusivity must be transparent."

That's an important distinction. If a team signs an exclusivity deal with a vendor and both sides announce they're working together, fans can evaluate that advantage. They can ask questions. They can pressure organizers to adjust. But if the contract is secret, if no one knows who has what, no one can evaluate. And what can't be evaluated can't be fixed.

I've seen this in football. When clubs began using analytics covertly, people didn't believe it was a problem. But when a few journalists began publishing lists of companies clubs partnered with, the story changed. Fans started asking questions. Leagues started reviewing. Regulation started appearing — slowly, unevenly, but it appeared.

What I want for esports is the same. Not a ban. A public list.

What fans should ask — and why they don't

Esports fans, across every market I've tracked, share one concerning trait: they know details about players, champions, patches, schedules, but they know very little about the commercial structure of teams and the tools they use.

That's not their fault. It's the fault of esports media organizations that have focused too much on intra-team drama and social media controversy, and too little on economic structure and governance. For years, I was part of the problem. I wrote hot takes about which teams should be criticized instead of pieces about who controls the data.

That's why I changed my focus.

Over the past three years, I've tried to write more about structure: how franchised leagues shift competitive advantage, how financial exclusivity deals work in football versus esports, how financial fair play rules change competitive structure. Those pieces got fewer readers. But they matter more.

I say what fans are afraid to hear, and they hate me for it. But this time, I'm not saying anything scary. I'm only saying something difficult: that your system is changing, and you haven't noticed.

Takeaway: A testable prediction

I don't predict the future, I excavate the past and throw it in your face.

The past here is: every time a technology becomes too important in a sport, that sport has to decide who the technology serves. Football needed nearly half a century for VAR and sixty years for financial fair play. Tennis needed two decades to unify its ranking system. Basketball needed thirty years to accept data analytics without treating it as unfair. Esports will be no exception.

My prediction: within three seasons, at least one major league — perhaps the LEC, perhaps the LCK — will announce regulations on third-party analytics tools. Those regulations won't ban entirely, won't permit entirely. They'll require public disclosure of vendor lists and a non-exclusivity commitment within competitive scope.

And when that happens, people will call it a governance step forward. But I'll call it what should have happened three years ago.

Not because I'm smarter than the organizers. Because I looked at the history of other sports, and I know how this story ends.

Tags

Jack Williams, iTero, GIANTX, AI coaching, AI coach, esports AI, League of Legends, LEC, Dota 2, teams, analytics tools, data exclusivity, Riot Games, Valve, esports governance, sports technology, sports news, esports tactics, esports machine learning, match report

Sources

Interview with Jack Williams on iTero, GIANTX, and the future of AI coaching in esports. Additional analysis based on public LEC records, Dota 2 The International 2026 history, and public match data from the 2026-2026 seasons.