Exclusive AI Deals in Esports: iTero, Giant X and the Unfinished Line of Cheating

I sat in front of the screen on a March night, watching the playoff round of a regional league. Between Game 3 and Game 4, the camera cut to the coaching area. A coach opened a laptop, typed rapidly, and closed it after exactly six minutes. His team entered Game 4 with a lineup that had never appeared all season. They won. I am not saying they cheated. I am only asking a simple question: who gave them that information, and how quickly? The six-minute break between games is becoming esports' new battleground. Not the battleground of players, but of software. And behind that software stand names like iTero, Giant X, and an unfinished debate about where the line between coaching and cheating lies. The story I want to dissect today revolves around Jack Williams and a tool called iTero. In a long-form interview, Williams discussed how iTero operates, his exclusive work with Giant X, and the likelihood of his product being copied. The interview has two clear sections: one on exclusivity and copying risk, and one on AI-assisted cheating. Both sit in the grey zone that tournament operators still hesitate to touch. Before going deeper, I need to be blunt about how I read pieces like this. Rumours are the surface. The system lies underneath. An interview about an AI tool is not a technology story. It is a story about power, about contracts, about who is permitted to know what and who knows it first. When an organisation like Giant X signs an exclusive deal with an analytics vendor, they are not just buying software. They are buying time. And in esports, time is the most expensive commodity. Let us start from the foundation. Professional esports has evolved through three preparation eras. Era one was intuitive: coaches reviewed VODs, took handwritten notes, relied on memory and experience. Era two was data-driven: teams hired analysts, built pick-ban databases, tracked individual statistics. Era three, the one we are entering, is machine learning: software does not merely store data but predicts, proposes lineups, and optimises decisions in real time. iTero sits in era three. That is why the debate is heating up. When software starts proposing, the question of responsibility and fairness stops being philosophical. It becomes legal and commercial. The first point I want to dissect is exclusivity. Williams discussed working exclusively with Giant X and the risk of being copied. This is a business structure very familiar in traditional sports. Major clubs sign exclusive deals with equipment suppliers, data centres, analytics firms. But in esports, this structure has a distinctive feature: the league operates within a closed ecosystem, where member teams cannot be relegated. Think about that. In a league with promotion and relegation, structural advantages tend to flatten over time, because weak teams get eliminated and strong teams get replaced. But in a closed franchise league, where every member holds a permanent slot, structural advantages are not competed away. They persist across seasons. If Giant X holds exclusive access to an advanced analytics tool, that advantage does not disappear after one season. It compounds. This is the point league operators will have to face, whether they want to or not. Historically, game publishers have intervened in similar structural advantages. They restricted coach communication during matches. They regulated break durations. They banned certain software from interfering with competition. Each time, the principle was the same: if a tool affects competitive outcomes, the operator has a duty to guarantee equal access. So why has the AI exclusivity contract not been touched? Three reasons. First, analytics tools do not directly interfere with the game. They do not change hitboxes, alter character stats, or affect competition servers. Technically, they fall outside the scope of current competitive rules. Existing esports rulebooks were written for a world where third-party software mainly means cheats. They have no language for third-party software as a tactical assistant. Second, measuring the competitive impact of an analytics tool is extraordinarily difficult. No one can prove that a specific pick-ban decision is the direct result of a recommendation from iTero. Coaches can always claim professional intuition. This is the blind spot in esports governance. Third, and perhaps most important, publishers have a commercial interest in non-intervention. If they acknowledge that AI tools can produce significant competitive advantages, they place themselves in the position of having to regulate a complex tooling market. Nobody wants to do that until forced. Now let us turn to the second section of the interview: AI-assisted cheating. This is where I think Williams framed the issue skilfully but also ambiguously. And that ambiguity is exactly the point. First, we must distinguish three time windows in competitive esports. Window one is pre-match: strategic preparation, opponent research, lineup construction. Window two is in-game: direct decisions during play. Window three is between games within a BO3 or BO5 series. Window two is heavily regulated. Every form of real-time in-game assistance is explicitly banned in every major title. There is nothing to debate there. So where does the real debate lie? It lies in window three. The interval between Game 1 and Game 2, between Game 2 and Game 3, is the largest grey zone in modern esports. Why? Because during that interval, coaches are permitted to talk to players. They are permitted to analyse. They are permitted to adjust strategy. The question is: what tools are they permitted to use, and how powerful may those tools be? iTero, if it operates as a between-game analytics tool, sits precisely in this grey zone. It does not violate current rules, because current rules do not prohibit out-of-game analytics software. But it changes the nature of the six-minute break. A coach without iTero enters the break with memory and notes. A coach with iTero enters with a model that has processed thousands of similar matches and proposes an optimised plan. This is not cheating under current definitions. But it is information asymmetry. And information asymmetry, in any sport, is the source of competitive advantage. I have watched many BO5 series over the past few years, and I have noticed an interesting pattern. Teams with strong analytics departments tend to win Games 3 and 4 at significantly higher rates than Game 1. This makes tactical sense: they have more data to adjust with. But if that gap keeps widening, the question becomes whether it reflects coaching skill or tooling. And this is where I must raise a factor the interview entirely omits: patch cadence. Each title has a different update rhythm. Dota 2 has large but infrequent updates. Major systemic patches arrive far apart, with long stable stretches between them. League of Legends updates far more rapidly, with biweekly patches. This difference directly affects the value of AI tools. In a stable-patch title, machine-learning models trained on historical data retain validity for long periods. They can predict accurately because the meta changes slowly. But in a biweekly-patch title, the half-life of any model is much shorter. AI's value shifts from solving the meta to detecting the new meta faster than opponents. This means a tool marketed identically across titles is hiding a truth: its value inverts depending on the ecosystem. In Dota 2, it is a knowledge advantage. In League of Legends, it is a speed advantage. Two products different in nature, potentially sold under one name. This is why I always distrust performance claims unaccompanied by evaluation methodology. Williams' interview provides no data, no sample size, no verification method. There is nothing wrong with that in a product-introduction interview. But as an analytical reader, I must flag it. Now let us discuss the most interesting part: copying risk. Williams mentioned the possibility of his product being copied. This is a legitimate concern for any tech company in a new market. But in esports, copying risk has a distinctive feature. The barrier to entry is not source code. It is data. A machine-learning model is only as good as the data that trains it. And high-quality esports data is not far away. It sits with teams, leagues, and game publishers. If a team like Giant X has exclusive access to iTero, it may also be supplying exclusive data to iTero. This is a closed loop: the more teams use it, the more data; the more data, the stronger the tool; the stronger the tool, the more teams want it. This means iTero does not merely compete on product quality. It competes on exclusive data access. And that is a far more defensible advantage than any algorithm. Anyone can write a machine-learning model. Not everyone holds a contract with a top-tier team. So why is Williams worried about copying? Perhaps because he understands that the data barrier holds only as long as the exclusivity contract is respected. And in a market without legal precedent, exclusivity contracts can be challenged in many ways. Another team could sign with a second vendor. A publisher could decide to release data publicly. A former employee could carry knowledge to a new company. This is where I want to pause and analyse more deeply, because it touches a larger structural question. Throughout esports history, sustainable competitive advantage has come from three sources: player talent, coaching quality, and organisational systems. AI coaching, if deployed correctly, can affect all three. It helps players understand the meta faster, helps coaches decide more accurately, and helps organisations build more standardised processes. But it can also erode one of those three sources: coaching quality. If a tool can propose optimal lineups, the value of an experienced coach declines. If a tool can predict enemy strategy, the value of a good analyst declines. This is a trade-off the industry has not discussed publicly. I have spoken with several semi-pro coaches in Korea over the past year, people working with lower-tier teams. They told me something worth pondering. They said analytics tools do not replace them, but they change the kind of work they do. They spend less time finding data and more time interpreting it. And to interpret data well, you need to understand the game more deeply, not more shallowly. This is an interesting paradox. The stronger the tool, the more human skill matters. Not because the tool is weak, but because the tool only makes suggestions. Deciding whether to trust that suggestion still belongs to a human. And under high competitive pressure, the decision to trust or distrust a model is the hardest decision of all. Now let us turn to the contrarian angle. This is the part most articles on AI coaching overlook. The official narrative about AI in esports is usually told in two directions. Direction one is technological: AI will revolutionise how teams prepare, help them understand the game more deeply, and raise competitive quality. Direction two is ethical: AI can be abused for cheating, and regulation is needed to prevent it. Both directions miss a question in the middle: the question of league fairness. Think carefully. If an AI tool is proven to create significant competitive advantage, and only some teams have access, the league operates on an uneven playing field. No one cheats. No one breaks the rules. But the structure of the competition has been distorted. This is what I call the governance blind spot. Tournament operators typically respond very quickly to cheating issues, because cheating is a clear violation. But they respond far more slowly to structural asymmetry issues, because no one is violating anything. There is no one to punish. Only a system operating in ways that favour some and disadvantage others. And in a closed franchise league, this problem does not resolve itself. It compounds across seasons. The advantaged team keeps its advantage. The disadvantaged team keeps falling behind. Over time, the gap becomes too large to close through ordinary effort. I am not saying Giant X is doing anything wrong. They are doing what every smart sports organisation should do: seeking competitive advantage. The problem is not them. The problem is a system that allows that advantage to exist without a balancing mechanism. And this is where I must raise another aspect the interview omits: differences between regional ecosystems. For years, I have watched how different regions approach technology in esports. Korea, with advanced internet infrastructure and gaming culture, tends to adopt analytics tools earlier and deeper. China, with large financial resources, tends to build in-house tools rather than buy externally. Europe and North America tend toward a market model, where third-party vendors compete to serve teams. Southeast Asia, including Vietnam, is often elsewhere. Teams here often lack resources to buy expensive tools, but they hold advantages in player numbers and adaptation speed. They start later but sometimes move faster. This creates an interesting dynamic. If AI tools become industry standard, resource-poor teams fall behind. But if those teams find creative substitutes, they may sustain competitiveness. Esports history is full of examples of resource-poor but tactically strong teams. This is why I think the AI coaching debate has an undervalued dimension: it may worsen regional inequality. Not because tools are designed for that, but because the cost of access reflects pre-existing resource gaps. Now let us return to the cheating question. Williams raised AI-assisted cheating as a topic in the interview. I think this is a skilful framing but an incomplete one. AI-assisted cheating, in the narrow sense, is using software to gain an impermissible competitive advantage. In the broad sense, it is using any tool to create an advantage opponents cannot access. The narrow definition is clear and enforceable. The broad definition is vague and hard to enforce. The real question is: which world do we want to live in? If we adopt the narrow definition, we keep rules clear but allow structural inequality to grow. If we adopt the broad definition, we address inequality but create a complex system with many new grey zones. I lean toward a third path: transparency. Instead of trying to define cheating in an AI context, leagues could require teams to disclose the tools they use. No bans. No punishments. Just disclosure. When information is public, fans and rival teams can judge for themselves. And public pressure is often more effective than any regulation. This is an idea I have not seen seriously discussed in any major league. It requires teams to give up information advantage, which no one wants to do voluntarily. But it may be the only path to sustaining esports competitiveness in the AI era. I have said rumours are the surface and the system lies underneath. In this case, the story of iTero and Giant X is the surface. The system underneath is how esports is struggling with the question of technological advantage in a competitive environment. And this is where I must confess something. I have been wrong before. In 2026, I wrote that esports teams would soon be forced to invest in data analytics to stay competitive. I thought it would happen within a year. It took three. I underestimated industry inertia, organisational slowness, and post-pandemic financial pressure. But in the end, that prediction held. I was just early. And that taught me a lesson about timing in analysis. A correct trend can still arrive later than you think. What matters is not abandoning that trend just because it is slow. So what comes next? I see three scenarios for the future of AI coaching in esports. Scenario one is integration. AI coaching becomes industry standard. Every major team uses it. Publishers develop their own tools and provide them to all teams. Inequality is minimised, but advantage shifts to teams that use tools better. This is the healthiest scenario but also the hardest to reach. Scenario two is fragmentation. Major teams build in-house tools. Smaller teams buy commercial tools. Differences in tool quality reflect differences in resources. Leagues struggle to manage growing inequality. This is the most likely scenario. Scenario three is regulation. Publishers decide to ban or strictly restrict third-party AI tools. They provide official tools with limited functionality. This is the most contentious scenario and could lead to litigation. In all three scenarios, the cheating question will keep haunting the industry. Because the line between coaching and cheating is not a straight line. It is a grey zone that moves with technology, with law, and with fan expectations. And this is the last thing I want to say about Williams' interview. He was right to place cheating and exclusivity side by side. These two topics seem distinct but are fundamentally one. Both concern the question: who is permitted to have an advantage, and is that advantage legitimate? In traditional sports, this question was resolved over decades with rules on doping, equipment, and sponsorship. In esports, we are only beginning. And we are beginning in an era when technology changes faster than our ability to write rules. That means we will have to live with ambiguity for a long time. Not because we are lazy, but because the speed of change exceeds the speed of governance. This is the defining challenge of esports over the next decade. And perhaps this is where I should end with a progressive thought rather than a summary. I do not think we will definitively resolve the question of the cheating line in AI coaching. I think we will learn to live with it, adjust it, and adapt to it. The way we have learned to live with every other controversial technology. What matters is that we do not stop asking questions. Do not stop testing. Do not stop demanding transparency. Because a system unchallenged will grow distorted over time. And in esports, where speed is everything, the time to challenge is always shorter than we think. Matches continue. Contracts get signed. Tools get updated. And somewhere in a dark arena, a coach is typing on a keyboard, waiting for an answer to a question perhaps no one is asking: did their victory tonight come from skill, from data, or from a contract no one read carefully?