iTero, GIANTX and the unanswered question: who is AI selling wins to in esports?

The third night of a Bo5 grand final, between game three and game four, there are only twelve minutes of rest. No one in the losing team's room talks about the opponent's composition structure. A coach opens a laptop, types a question into a dialogue box, and lets a machine-learning model answer: "Why did we lose the fight at minute 24?" Twelve minutes later, that team wins game four, then wins the whole series. I have rewound that tape no fewer than seven times over two years, and every time I rewind it, an old question resurfaces: we are arguing about whether AI should step into the competitive room, while it has been sitting there for a long time — just nobody dared name it. The shock does not come from the scoreline, but from the place we refuse to look. The recent interview with Jack Williams about iTero and GIANTX that I read touched exactly that spot. It does not tell the story of a victory. It tells the story of who owns the tool that manufactures victories — and that is a far more uncomfortable story. CONTEXT: A NICHE INTERVIEW AND TWO NAMES WORTH NOTING The name Jack Williams appeared in a niche B2B interview, the kind mainstream audiences rarely encounter. The subject revolves around iTero, an AI-driven data-analytics product built to assist coaching, and GIANTX, an esports organisation believed by industry insiders to be EMEA-based, formed after the merger of Excel Esports and Giants Gaming, with a foothold in the professional League of Legends ecosystem. The piece revealed two subheadings: one about an exclusive partnership with GIANTX and the likelihood of being copied, another about AI-assisted cheating. Those two frames — commercial and integrity — sit on either side of a gap the article never touches: league fairness. That is the thing worth discussing. I am not an insider at iTero or GIANTX. I am a contrarian journalist who has sat long enough in press rooms and backstage corridors to notice a rule: every time a new tool enters esports, the industry goes through exactly three phases. Phase one, silence — nobody admits to using it. Phase two, one team openly uses it, usually the team that is winning. Phase three, everyone else splits into two camps: the copycats and the moralisers. iTero and GIANTX sit between phase two and three, which is why this interview is more worth reading than it appears. Over the past three years, AI has crept into every corner of professional esports. It did not invade through a thunderous event. It invaded through small things: a tool predicting win probability after a dragon steal, a pick-ban suggestion system built on head-to-head history, a coaching assistant that answers in three seconds instead of three hours. And because it invaded quietly, nobody noticed the moment it became indispensable. This is a major-tournament season. Major-tournament seasons do something few notice: they compress every emotion, every dispute, every question about fairness into a single line — who gets to use what? And AI, in this story, is precisely the answer to that question, hanging in midair over the stage. ANALYSIS: AI IN ESPORTS DOES NOT LIVE IN ONE PLACE The first thing to clarify is that AI in professional esports does not live in one place. It lives on three tiers, and each tier carries a different level of controversy. The first tier is post-match analysis. Rewatching tapes, finding patterns, building models of opponents. This tier is almost uncontroversial, because every major team has had a manual data-analysis crew for years, and AI is merely a faster, cheaper, more accurate way of doing the work. Nobody calls an analyst a cheater, so AI replacing part of an analyst's job is hard to call cheating either. The second tier is pre-match and between-game support. Predicting picks and bans, calculating probabilities, suggesting adjustments after losing a game. This is the genuine grey zone. Between games in a Bo5 series, the break lasts only ten to fifteen minutes. The question of "what a coach is allowed to bring into that window" has no uniform answer across venues. Some tournaments ban paper. Some allow laptops but ban internet. Some say nothing clear at all, and that silence is a gap. The third tier is real-time support — intervention while the match is running. This tier is banned in every major tournament, and because it is banned, it is no longer an interesting debate. Everyone agrees this is red territory. So the real debate lives on the second tier. And that is where iTero, if its product truly does what the name implies, would generate the most value. It is also the most dangerous spot, because it sits exactly on the line between "a tool that helps a team play better" and "a tool that lets a team know in advance what the opponent will do". There is a technical variable most public debate skips: the patch cadence of each title determines what AI is actually worth. In Dota 2, Valve patches infrequently and heavily. A major update can change nearly the entire way the game is played, but between two major patches lie long stretches of stability. That rhythm favours statistical and machine-learning tools, because a model trained on historical data retains its value for a long time, even across months. In League of Legends, Riot patches every two weeks. That fast rhythm shortens the lifespan of every learned pattern. There, AI's value shifts from "solving the meta" to "detecting the meta shift faster than the opponent" — a tempo advantage, not a knowledge advantage. You do not need to know the right answer. You need to know when the right answer changed. This is the point the industry rarely admits. People say "AI coaching support" as if it were a single product sold to a single market. In reality, one product marketed identically for both Dota 2 and League of Legends is a red flag. The two ecosystems reward two different kinds of intelligence. Selling the same thing to both means either the product is shallow, or the seller is diluting the definition to suit every palate. Now the commercial side, where the story becomes more concrete. The article discusses iTero's exclusive partnership with GIANTX and the likelihood of being copied. This is the classic business model of every B2B tech sector: sell privately to one large client first, use that client's reputation to sell to the rest of the market, and use that first client as a proof of concept. The problem is that professional esports is not an ordinary market. It is a closed market, where every rival competes on the same stage, before the same audience, under the same rulebook. When an exclusive tool lands in one team's hands, the thing affected is not only the vendor's profit. The thing affected is the result on stage. Consider the structure of a closed franchised league, the LEC model. Every member is a permanent member, nobody is relegated, nobody is promoted. That means a structural advantage held by one member persists across seasons, instead of being competed away as in an open system. In an open league, other teams can be promoted, re-equip themselves, and close the gap. In a closed league, that gap freezes. This is why an exclusive deal in a franchised league carries far more structural weight than the same deal in an open system. And this is where the publisher itself must be questioned. If a tool truly affects competitive outcomes, the league operator will soon face a fork: either force equal access for all teams, or restrict the tool. History gives us a clear precedent. In-match communication between coaches and players was regulated step by step: from fully permitted, to limited in number and content, to nearly closed in many phases. Every tightening step carried the same reason: protecting the integrity of the competition. Analytics tools will travel that exact road, only slower, because they are more invisible. Nobody sees the AI in a coach's headset the way they see a microphone. Then comes the word "copied". In this industry, copying is a ritual. One team wins, the rest imitate. But when the thing copied is a digital tool rather than a tactic, the story changes. An AI model can be retrained on another team's data, but it cannot copy the intuition forged over thousands of hours between a coach and that model. What rivals can copy is the interface. What they cannot copy is the process. And in most cases, people sell the interface while keeping the process. That is why "anti-copy" promises in AI deals often sound better than they are. The integrity frame — AI-assisted cheating — deserves more attention and is easier to misread than either. Since generative AI emerged, esports has feared one scenario: a player using AI during a match to make real-time optimal decisions. That fear is somewhat misplaced. In live competition, every signal is monitored: screens are recorded, headsets are controlled, referees stand behind players, and the latency between a question and a useful answer during a three-second teamfight is too large for AI to intervene undetected. The real-time cheating window is nearly sealed. The window that is genuinely open sits between games, in that twelve-minute break nobody dares name. There, a coach can ask AI anything, and no rule defines it as cheating. That is the central paradox of the whole story: we guard the competitive room down to the centimetre, yet leave the backstage corridor wide open. CONTRARIAN VIEW: THE QUESTION IS ASKED IN THE WRONG PLACE I know I am going against a popular belief. The majority holds that the problem with AI in esports is integrity — how to catch the cheater. I hold that the larger problem lies in structural fairness, and it cannot be solved by catching anyone. Try a comparison. If tomorrow Riot announced that every team in the LEC would be granted access to an analytics tool like iTero, we would cheer "fairness". But would it be fair? One team has three analysts, another has one. One team has a coach who knows how to ask the right question, another has a coach who only knows how to read reports printed by the tool. Equal access to a tool does not mean equal ability to use it. That is the biggest blind spot in every debate about AI in sport, not just esports. Point two: the question "should AI exist in esports" is already obsolete. It exists. A team that does not use it will lose to a team that does, and that will happen before any ethics board can convene. The real question is not "should it", but "who is allowed, and who decides". In a major-tournament season, when every eye turns to a handful of national teams and a handful of giants, the one allowed is the one who can afford the tool, and the one who decides is the publisher — a party with its own interest in staying quiet. Point three: we talk about AI as if it were a neutral entity. But an AI tool trained on one region's data carries that region's bias. A model that learned from thousands of European matches will suggest moves that are rational the way Europe plays. Handing it to an Asian team without warning is a form of tactical-cultural imposition. I have written about such stories in football, and I believe they will recur in esports, only under a technical veneer. I also have to say this: I may be wrong. Perhaps iTero is simply a wholesome post-match tool, and its exclusivity with GIANTX is a small deal not worth discussing. Public evidence is insufficient to assert anything more serious. But if I am right, this industry is entering a zone where its rules are written by commercial contracts rather than competitive regulation — and the history of every sport shows that whenever power leaves the pitch to enter the boardroom, the fans are the last to find out. WHAT REMAINS An article that provokes a boycott is an article touching someone. I do not want this piece to boycott anyone. I want it to force this industry to write down a definition. Naming a sport is how I remind it to look at itself. And esports, at twenty-five years old, needs a mirror more than a leaderboard. The question I leave readers is very concrete: if your team loses a Bo5 because the opponent had a better AI tool, what would you call it — talent, money, or cheating? And who do you think should answer that question? I leave the conclusion to the reader. As for me, I will keep watching that twelve-minute tape, because I believe the most important thing in sport is not who wins, but whether the rules of the game are still worth playing by.