Jack Williams, iTero, and the Regulatory Vacuum of AI Coaching in Esports

In August 2026, at Gamescom in Cologne, Natus Vincere lifted the Aegis of Champions - the first The International title in Dota 2 history. At the time, a British teenager named Jack Williams was watching on a small monitor, jotting down every tower push in a paper notebook. He could not have known that fourteen years later, he would become the central figure in a debate no tournament has dared to define clearly: the legal and ethical boundary of AI-powered coaching software in professional competition. Williams's recent interview revolves around three axes: iTero - the analytics platform he works with, his exclusive relationship with the organisation GIANTX, and his fear of being copied. The conversation appears to be about business, but when I reached the section on AI-assisted cheating, I realised the real issue lay elsewhere - in the gap between competitive rules and technological reality. Jack Williams runs iTero, an analytics platform that applies machine learning models to support the preparation process of professional esports teams. Unlike traditional statistical tools that only aggregate post-match metrics, iTero aims to model opponents, predict pick/ban trends, and optimise practice plans based on historical data. iTero's exclusive partner in Europe is GIANTX - an organisation formed through the merger of Excel Esports and Giants Gaming, currently competing in the League of Legends EMEA Championship (LEC). This is a closed-franchise league, meaning member teams are not relegated on a seasonal basis but hold fixed slots. With that structure, an advantage - such as exclusive access to an advanced analytics tool - persists across seasons rather than being equalised by competitive pressure. Williams admits in the interview that he worries his product will be copied. This reflects a familiar paradox: the value of an analytics tool lies in its exclusivity, but that very exclusivity makes it a target for imitation. The section on AI-assisted cheating shows that even tool developers lack a clear answer on usage limits. One point must be stressed: the interview provides no patch numbers, balance versions, or specific pick/ban data. This is a B2B thought-leadership text, not a tournament report. Any assessment of iTero's real-world effectiveness in a specific season is therefore impossible from the available source. The meta actually shifting here is not inside the game but at the layer of preparation tooling. Esports does not need luck, it needs people who read the meta faster than the server. That is true for players, but twice as true for software developers selling meta-reading as a service. Patch cadence explains why the value of AI coaching inverts across titles. In Dota 2, Valve maintains infrequent, structurally disruptive major patches, with long stretches of stability in between. Machine learning models trained on historical data retain their validity over a wider time window, favouring deep statistical tooling. By contrast, League of Legends patches every two weeks, shortening the half-life of any learned pattern. There, AI's value is no longer decoding the meta but detecting the meta delta faster than opponents - a tempo advantage, not a knowledge advantage. This matters for iTero. If Williams's product is marketed identically across both ecosystems, that is a red flag. A tool optimised for slow patch cadence is useless in a fast-patch environment, and vice versa. Without title-specific adaptation, the competitive edge is just a marketing story. Two concepts being lumped together must be distinguished. Pre-match, between-game and post-match assistance is the grey zone under debate. Real-time assistance is already explicitly prohibited in every major title, leaving nothing to argue about. The genuinely interesting grey zone lies in the between-game window of BO3 and BO5 series - a period current rules have not clearly defined on who may use what tool. The second layer is the exclusivity problem in a closed league. In the LEC, all members are permanent members, with no relegation pressure. A team's structural advantage - such as exclusive access to iTero - persists across seasons rather than being competed away. In an open system with promotion and relegation, that advantage self-equalises over time because weak teams get eliminated and strong teams imitate. In a closed system, it accumulates. Every season is a ritual, and the analyst is merely the scribe of its omens. Here, the omen is institutionalised asymmetry. Notably, the league operator - in this case Riot Games - faces a familiar governance situation. If a tool materially affects competitive outcomes, the operator will face pressure either to mandate equal access or to restrict the tool. This is precisely the path in-game coach communication travelled - from freedom to step-by-step control. The cancelled Seoul derby of 2026 was a test for every prediction algorithm, and the lesson remains intact: when an external variable changes, old models collapse. With AI coaching, the external variable is the unwritten rulebook. The third layer is the addressable market structure by publisher. Valve and Riot Games have long held differing views on the permissiveness of third-party tooling. If so, AI coaching vendors face entirely different addressable markets per title. A product that succeeds in Dota 2 may be illegal in League of Legends, not for technical reasons but for legal ones. The central paradox the interview does not resolve: Williams sells exclusivity, but exclusivity in a loosely regulated ecosystem is an asset, while in a tightly regulated one it is a risk. The more successful the product, the more likely it is to be regulated. The more exclusive, the more likely it becomes a target of anti-cheating rules. Between the transfer figures lies a story no report records. Here, that story is AI coaching being priced as a technology product while its essence is an unresolved governance question. No report records that iTero's value depends on whether Riot Games defines the boundary clearly. If Riot bans it, the entire business model evaporates overnight. The popular view holds that this is a technology story - about AI changing how teams prepare. That is true but insufficient. The contrarian view: this is primarily a story about unwritten rules, and every claim about technical effectiveness is inference from a text with no data. The interview carries two headings: working exclusively with GIANTX and the likelihood of being copied; and AI-assisted cheating. The first frame is commercial. The second is integrity. Between them sits an unfilled third frame: league fairness. This is the least explored angle, and it is the one that determines iTero's long-term value. I do not believe in intuition; I believe in numbers that speak after being asked the right question. Here there is no number to ask - no sample size, no evaluation methodology, no before-and-after comparison data. Every effectiveness claim is unverifiable from the available source. Beware also of false correlation. GIANTX signing exclusively with iTero does not mean their results will improve. That mistake years ago taught me that data never lies, only the reading does. The correlation between tool use and better results may stem from a bigger budget, a stronger roster, or simply luck in a short window. No control group, no conclusion. The next cycle will be shaped by a line in the tournament rulebook rather than by a new patch. If Riot Games clearly defines the limits of AI coaching tools next season, this entire market will be re-priced. The question is not whether iTero works. The question is: when you sell a tool that can change match outcomes to a single team in a closed league, who decides whether that is fair?