AI Coaching in Esports and the iTero–GIANTX Exclusive Deal: A Governance Boundary Not Yet Written into Rules

September 2026, Incheon. I reopened a spreadsheet I first built in 2026 and found it obsolete in an uncomfortable way. That spreadsheet priced a player by the growth rate of their social media following, added to a performance efficiency index and a commercial conversion coefficient I invented after many nights of trial and error. When Jack Williams of GIANTX spoke about his team using iTero's artificial intelligence tool as an exclusive advantage, the new variable was not the player. It was the software that player was permitted to touch. An exclusive contract inside a closed league, where nobody is relegated, does not create a sprint advantage. It creates an accumulated advantage across seasons. And accumulated advantage in sport is how empires are built, or how a league is quietly strangled before anyone notices. The interview I read had two clearly marked headings. The first dealt with working exclusively with GIANTX and the likelihood of being copied. The second dealt with AI-assisted cheating. Both frames sound reasonable: the commercial frame and the integrity frame. Between them sits a third frame that has been left empty, and that frame is where the real money flows. League fairness. This article exists to fill that gap. Before going deeper, one methodological point must be stated plainly. The source material for this story is alarmingly thin. Of thirteen information points I could reach, ten describe the biography of the original article's author rather than the subject of the interview. Only three points actually carry content about Jack Williams, iTero and GIANTX, and of those three, two are sourced only to headings rather than body text. That means anyone telling you they know exactly what iTero sells, at what price, for how long, is fabricating. Including me. What I can do is analyse the structure of the deal based on entity names, two section headings, and what this industry has repeatedly done in the past. Context: three layers of power and one new tool. To read the iTero-GIANTX story as a governance document rather than a technology story, it must be placed inside esports' three-layer power structure. The first layer is the publisher. Valve and Riot Games hold the right to determine the rules of play, the schedule, and most importantly, which tools may appear inside the competitive area. This is the least questioned layer because fans rarely read terms of service. The second layer is the league. The franchise model, such as the LEC, turns teams from entities that must fight to survive into fixed members, free of relegation pressure, without an automatic mechanism of merit-based elimination. The third layer is the team, where sponsorship deals, player contracts, and now analytics tool contracts are signed. An AI coaching tool enters all three layers simultaneously. It is a commercial product at layer three, but its legal existence is decided at layer one, and its competitive effect appears at layer two. Something that touches all three layers of power while no dedicated body of law exists for it is the definition of a governance gap. In my nineteen years observing this industry, from my role as an esports athlete in 2026 to club financial analysis in Korea, every time a new tool appeared inside the competitive area, the industry reacted in exactly one sequence: denial, then a temporary ban, then conditional legalisation, then finally writing it into rules when it was already too late to correct the error. In-game coach communication went through exactly that sequence. Post-match data analysis did too. A note on the story's date. The original article does not state its date in the material I could reach. But it contains a calculable temporal anchor. It says Natus Vincere lifted the Aegis of Champions at Gamescom fourteen years ago. Natus Vincere first won The International at Gamescom in 2026. Simple subtraction places the article around 2026. This is arithmetic inference from the article's own wording, not speculation. It matters because any analysis of esports regulation must be tied to a date, and in esports, two years is the distance between two generations of rules. The second notable point is that the Na'Vi detail sits in the author's biography, not in the interview body. It is nostalgic memory, not competitive data. Anyone reading that detail as a signal about the current Dota 2 landscape is committing a category error. A 2026 story says nothing about a 2026 patch. Core analysis: what iTero actually sells. iTero does not sell an artificial intelligence model. The model is something anyone can buy or build. What iTero sells is a data access window and a speed of patch decoding faster than opponents. Separate these two variables. The first is the data window. In professional esports, match data is not opened publicly to everyone at the same time. Publishers control who accesses what data, when, and through which channel. A tool with an official partnership relationship can touch data at a deeper, earlier, or rawer layer than a self-built tool. That gap is not a gap in algorithms. It is a gap in access rights. And access rights are a commercial asset, not a technical achievement. The second variable is patch decoding speed. This is where the story becomes interesting, because the value of an AI tool differs across titles. It depends directly on the publisher's patch cadence. Valve, with Dota 2, has an infrequent and disruptive patch cadence: large systemic patches, then long stable stretches between them. In this rhythm, a machine learning model trained on historical data retains validity for a longer window. Advantage belongs to depth of historical modelling. Whoever has more historical data, whoever processes it more deeply, wins. Riot Games, with League of Legends, has a dense patch cadence on a two-week cycle. In this rhythm, any pattern learned from data has a short lifespan. The value of an AI tool shifts from decoding the patch to detecting the patch delta faster than opponents. This is a tempo advantage, not a knowledge advantage. Whoever detects which direction the patch is shifting one round earlier, wins. These two titles demand two opposing product philosophies. A product marketed identically across both titles is a warning sign. If iTero claims the same value in Dota 2 and League of Legends, either they do not understand the difference, or they are selling something generic into two markets with entirely different temporal structures. I once built three parallel valuation models back when I worked at Incheon United, and the biggest lesson was not which model was right. The lesson was that each model is right in one context and wrong in another. Every valuation model is wrong. The question is in whose favour it is wrong. The same logic applies to AI tools: their value depends on where they are wrong, and who pays for that wrongness. Core analysis: the economics of an exclusive deal. The first section heading deals with working exclusively with GIANTX and the likelihood of being copied. Read the phrase likelihood of being copied carefully. The interviewee speaks of the risk of opponents imitating. But an AI tool, by its nature, can be copied in two very different ways. The first is product copying. A competitor builds a similar tool with a similar algorithm. This is the ordinary risk of any software product, and it has nothing to do with esports. The second is outcome copying. A competitor does not build a tool, but observes how GIANTX plays, how it bans and picks champions, how it rotates lineups between games, and reverse-engineers what the tool is advising. In esports, every behaviour on the field is public. A tool, however secret, leaves traces through the decisions of whoever uses it. The second risk is the real risk. And it cannot be solved by an exclusive contract. It can only be solved by making the tool generate an advantage that is unobservable from outside, which is nearly impossible in a discipline where every match is streamed live and every decision is re-analysed. In other words, an exclusive contract buys GIANTX a window of time, not a permanent advantage. The correct commercial question is not whether iTero gets copied, but how long the exclusivity window lasts and what iTero's residual value is after it ends. In software, every exclusive advantage has an expiry date. In esports, that expiry arrives faster because the velocity of tactical knowledge diffusion is higher than in any other industry. Core analysis: asymmetry in a closed league. This is the most important part, and the part both of the article's headings omit. GIANTX, by my understanding of the industry context, is an organisation rooted in the EMEA region competing in the LEC system, formed from the merger of Excel Esports and Giants Gaming. If that is accurate, the regulatory framework governing the iTero deal is Riot Games' third-party software and competitive integrity rules. I must stress this is my background industry knowledge, requiring verification, and I should also flag the possibility that this is a different entity rendered as Giant X. If GIANTX truly operates in a franchise league, the structural asymmetry here is severe. In an open league, where teams are promoted and relegated by results, a structural advantage is competed away over time. Weaker teams learn, imitate, overtake, or are removed from the league. The natural elimination mechanism acts as a safety valve. In a franchise league, that safety valve does not exist. Members are fixed. A structural advantage held by one member is not competed away. It accumulates. Season after season, the team with the better tool is more likely to win, attracts more sponsorship, signs better players, and reinforces its advantage. This is a positive feedback loop, and inside a closed league, it has no natural stopping point. Numbers must be placed in context. A marginal competitive advantage in sport is worth not in itself, but in its effect on result distribution. If an analytics tool raises a team's win probability from fifty percent to fifty-two percent per match, then across a long season the gap with the rest is no longer two percent. It is an entire difference in standings, in international qualification slots, in sponsorship revenue, in brand value. In sport, a small margin multiplied by a long time horizon creates a large gap. That is the basic mathematics of accumulated advantage, and it is why asymmetry in a closed league is more severe than asymmetry in an open league. Why the commercial and integrity frames miss this. The commercial frame asks: can iTero hold its advantage, will it be copied, will it make money. That is the investor's and the vendor's question. The integrity frame asks: is using AI cheating. That is the referee's and the community's question. Both frames skip the league's question: whether an exclusive arrangement creates an uneven playing field in a competition where nobody is eliminated for losing. This is not an abstract ethical question. It is a concrete governance question that every franchise league in the world will have to answer within a few years. How it is answered will determine not only iTero's future but the future of the entire esports analytics industry. Look at how traditional sports leagues handled a similar problem. When data analytics technology became widespread in football, leagues did not ban it. They standardised access. They required every team to have equal access to data, and prohibited sharing exclusive data from a vendor with a special relationship to one team. The principle here is clear: technology is free, data must be fair. Esports has no such principle yet. And while esports lacks it, a deal like iTero-GIANTX can exist without violating any rule, simply because the rule has not been written. Esports is a mirror, not a rival. There is a misreading I see repeated in conversations about AI and sport. People set esports against football, as if these were two worlds with two sets of laws. I do not think so. Esports is not football's rival. It is the mirror exposing the entire spending habit of the sports industry. What is happening with iTero and GIANTX will happen to football within five to seven years, just one beat slower. Football already has exclusive data deals between clubs and analytics vendors. Football already has disputes over real-time data access. Football already has clubs paying analytics companies for an advantage their opponents lack. The only difference is that football does it more slowly, with more intermediary layers, and with a thicker governance system shielding some of the asymmetry. Esports does the same thing faster, more directly, and with fewer intermediary layers. That is why anyone interested in sport should read this story. It is a public laboratory for a problem that will spread to every discipline in the coming decade. I have seen this once, at a much smaller scale. In 2026, when the pandemic emptied stadiums, I organised a brainstorming session with six marketing staff and proposed four new revenue models. Two failed. The remaining two, including virtual advertising on broadcast, brought in significant revenue within three months. What I learned was not which model won. What I learned was that crisis does not create new problems. It only exposes models that were long dead. Crisis does not destroy. It defines. Apply that logic here: the AI coaching wave does not create a new fairness problem. It only exposes that franchise leagues already had an asymmetric structure, and that structure was never tested because no tool was ever powerful enough to amplify it. The contrarian angle: the problem is not AI, it is exclusivity. The popular conclusion I predict will spread through the community is: AI is ruining esports fairness, AI must be banned, we must return to the era of purely human coaching. I think that conclusion is wrong in where it places the emphasis. Public AI does not ruin any league. If every team has access to the same AI tool, or to AI tools of equivalent quality, that is merely a general upgrade of the whole skill floor. It is like the spread of video analysis: everyone uses it, the floor rises, and nobody gains a structural advantage from using it. What ruins a league is exclusivity. One team having access to a tool others do not. Data open to one and closed to another. Different access speeds among members of the same league. In other words, the real integrity problem of the coming decade is not the algorithm. It is the contract. This leads to a specific prediction. When leagues begin writing rules on AI coaching, they will not write rules about algorithms. Writing rules about algorithms is impossible, because algorithms change faster than rules. They will write rules about data access, about tool fairness standards, about the obligation to share analytical advantage among members. They will do to AI what they did to broadcast rights and data rights: standardise at the data layer, liberalise at the tool layer. And when that happens, iTero's business model will shift from selling exclusivity to selling a standard. From selling a private advantage to one team to selling shared infrastructure to a whole league. That is an entirely different commercial transition, and it demands a different kind of relationship with the publisher. If I am right, iTero's long-term value is not in the exclusive contract with GIANTX. It is in iTero's position within the future rulebook that Riot Games and Valve will write. A club does not need a full stadium to make money. It needs to know what an empty stadium is saying. What nobody has said: the risk sits with the publisher, not the team. There is a third party in this story that neither the commercial nor the integrity frame mentions. The publisher. If a league allows an exclusive tool deal to exist, the publisher is carrying two risks at once. The first is competitive risk: the league becomes less attractive as results become more predictable. The second is legal risk: an exclusive arrangement in a closed league may be examined through the lens of anti-competitive conduct, especially in markets with strict monopoly frameworks. The second risk is rarely mentioned but may be the largest. For years, major esports leagues have faced questions about their franchise structures and their competitiveness. An exclusive tool deal, viewed through that lens, could turn a technology story into a serious governance problem. This is why I believe publishers will react sooner than the community expects. Not out of ethics. Out of risk. As I say this, I recall a debate from my time working at the Korea Football Association around the 2026 Russia World Cup. Back then, I was tracking sponsorship effectiveness and found a paradox: the Korea versus Mexico match on 23 June 2026 drew millions of online viewers, but jersey sales revenue fell year on year. I argued that the traditional broadcast licensing model was missing significant revenue from digital platforms. I was opposed. But the lesson was not that I was right. The lesson was that large organisations do not react to truth. They react to risk. When risk is not yet present, truth can wait. When risk is present, truth becomes urgent. That is why I follow this story at the publisher layer, not the team layer. How to read the numbers correctly. I am someone who hunts for hidden value in markets, and that means I am especially suspicious of impressive numbers. In this story, no financial figure is disclosed. No contract value, no duration, no scope. The two section headings discuss exclusivity and cheating, not money. That is normal for a thought-leadership interview, but it means any financial analysis of this deal is currently speculation. I do not want to write an article in which I pretend to know numbers I do not know. Throughout my career, my principle has been to question every number, including numbers from official reports. But questioning a number is different from inventing a replacement number. What I can do is point out four contexts into which the missing number must be placed. The first is the revenue structure of the esports analytics industry, which is small relative to total esports revenue but carries much higher margins. The second is the short customer lifecycle in this industry, where a team can terminate a contract after each season. The third is the cross-effect of an exclusive contract on a team's other contracts. The fourth is the option value of becoming the standard vendor if a league standardises data access, a scenario that could carry far more value than any exclusive deal. All four contexts point to the same thing: the value of this deal is not in the amount of one contract. It is in the position that deal secures within the future rulebook. Takeaway: watch the rules, not the product. Esports fans are being invited to watch a product demonstration. AI demos, accuracy claims, promises of faster and deeper analysis. These are easy to see, easy to be drawn to, and easy to forget. What matters more sits where nobody shines a light. The league rulebook, the third-party software clause, the data access conditions that Riot Games and Valve will publish within two or three years. Those documents will decide who holds an advantage, who is left behind, and which structural asymmetries will be legalised in esports. In a franchise league, every exclusive advantage is an accumulated advantage, and every accumulated advantage is an advantage borrowed from the league's future. When fans sit down to watch the 2026 season, they may be watching a race decided before the opening whistle, by a contract nobody in the stands can read. What I will be watching is not the next version of iTero. I will be watching Riot Games' rulebook page. There, in a line of terms almost nobody quotes, the competitive fairness of esports is being written, or quietly overlooked.