Jack Williams, iTero and GIANTX: When AI Becomes Coach, Is the Esports Arena Still Fair?

A small detail sits in the author biography section of the interview: Ollie hopes to one day recreate the moment Natus Vincere lifted the Aegis of Champions at Gamescom 14 years ago. If counted from The International 2026 championship, that number places the article around 2026. What matters is not the writer's memory — but that it is the only component, across all extractable content, that directly references a specific tournament. The rest of the conversation with Jack Williams — centred on iTero, GIANTX, and the future of AI-powered coaching tools — contains no bracket, no patch, no roster, no play. No balance statistics. No win rates. No match data. Ten of the article's thirteen information points describe the author rather than the subject. Only three touch the real topic, and two of those are drawn from section headings, not body text. That is a signal. When an esports interview cannot contain a single moment of competition, its true subject is not competition. It is governance. It is commerce. And it is an unnamed battle being fought in meeting rooms rather than on screens. I have sat backstage at enough tournaments to know that the biggest shifts in this discipline rarely come from a patch. They come from a contract clause. And this time, that clause has a name: exclusivity. For more than a decade, match analysis in professional esports happened in two ways: with the human eye and with spreadsheets. Coaches and analysts sat through hundreds of hours of footage, timestamping manually, drawing movement maps by hand, noting every vision control event. All of it depended on a scarce resource: human time. Around 2026 to 2026, a new class of tools began to appear. They did not play the game. They read replays, extracted thousands of events per match, grouped them into patterns, and returned to coaching staffs a list of anomalies: a vision control lapse repeated by the same rule, a teamfight zone with an unusually high loss rate, a time window where a team loses map control in an identical pattern across matches. iTero sits in that group. Jack Williams, the interviewee, represents a software company trying to sell its product to professional organisations. But software in esports has never been just software. Because what teams buy is not a tool. They buy advantage. And advantage, inside a closed league, is a limited commodity. That is why the iTero and GIANTX arrangement became the centre of the conversation. As the article phrases it, iTero partners exclusively with GIANTX. The word exclusive is not decorative. It is the essence of the entire story. GIANTX — an organisation reportedly formed through the merger of Excel Esports and Giants Gaming, competing in the EMEA regional league system — holds access to a tool its direct rivals do not. In an open circuit, that is not especially serious. Teams are promoted and relegated constantly, advantages dilute over time, and the market finds ways to copy or replace the tool. But in a closed league — where every member is a permanent member, where there is no relegation slot to punish stagnation — a structural advantage is never culled. It accumulates. It persists season after season, year after year. That is the context the interview does not state but places its finger directly upon. Two disclosed headings — one on exclusive partnership with GIANTX and the likelihood of being copied, one on AI-assisted cheating — form the two ends of a taut string. One end is a commercial question: how do you protect competitive advantage when your product can be mimicked? The other is an ethical question: where is the line between a legitimate support tool and a form of cheating? Between those ends sits a third framing no party wants to name: should a league allow one member to hold an analytical advantage another does not? To understand why AI coaching has become a hot topic, it must be separated from any specific game. The tool itself does not depend on a title. It depends on data. And the speed at which data changes differs entirely across disciplines. In a slow-patch title like Dota 2, where Valve releases major patches at long intervals and lets the meta stabilise for months, a machine-learning model trained on historical data retains its value for extended windows. The advantage of AI here is depth: the ability to model history, to recognise tactical patterns that survive across versions. In a fortnightly-patch title like League of Legends, the half-life of any behavioural pattern is compressed. Here, AI's value shifts from cracking the meta to detecting meta drift faster than opponents. That is a tempo advantage, not a knowledge advantage. This is the core distinction most public discussion skips. A tool marketed identically across both disciplines is a red flag. Its value-creation logic inverts between the two environments. Selling the same product for a title that needs deep historical modelling and one that needs a real-time drift radar means you are selling two different things under one name. If the interview does not draw that distinction, then either Jack Williams is discussing a specific title the article does not name, or the product's value claim should be read carefully. That is why I believe the real debate in this interview is not whether AI can be a coach. It lies in the question of permitted space. AI-assisted cheating — the subject of one disclosed section — almost certainly concerns the between-game window, not in-game. Real-time assistance during play has long been unambiguously banned in every major title; there is nothing to debate. But the window between game one and game two of a best-of-three, when coaching staff leave the stage and sit in the war room, is an undefined grey zone. If an AI tool reads data from the game just finished and offers tactical suggestions for the next, how is that different from a human analyst working at extraordinary speed? And if the answer is that it is no different, why should one team owning that tool be considered a problem? The answer lies in the word exclusive, not in the word AI. One point about the nature of data in esports should be made clear. In most major titles, match data is a shared asset in a certain sense — generated in official matches, witnessed by the public, archived under the organiser's rules. But the layer of data AI tools create is different. It is second-order data: analytical output, modelling, prediction. It does not belong to the league. It belongs to the software company. And the software company may sell it to whomever it wishes. This is the point current regulations in every major league have yet to touch. Rules ban players from using cheating software during play. Rules specify when coaches may communicate with players. But rules do not yet specify who owns the analytical output of public data. That is the legal gap — and it is the commercial gap. There is a subtle difference between the two biggest publishers worth putting on the table. Valve and Riot Games pursue two distinct philosophies on permitting third-party tooling. If that difference is real, an AI coaching vendor faces two markets of entirely different size depending on which publisher controls the ecosystem. A tool may be legal and prevalent in one discipline but viewed as a threat in another. This analysis needs verification against each publisher's current policy, not recollection or rumour. I witnessed a similar gap a few years ago, during the months when leagues had to compete in empty stadiums. When all physical interaction vanished, people assumed the discipline would lose its vitality. Reality proved the opposite. On a summer final night I attended, when the home side completed a reverse sweep in the last game, thousands of people screamed together in a virtual chat room — while the real stadium held not a single soul. The silence of physical space could not kill the resonance of emotion. I bring that up because it is directly relevant. Gaps do not fill themselves. Someone fills them. In the empty-stadium case, fans filled it with their voices. In the case of the regulatory gap on AI, a software company is filling it with an exclusivity clause. The popular narrative about AI in esports splits the world into two camps. One camp proclaims AI will democratise analysis, handing the power of big teams down to every small organisation. The other warns AI will destroy competitive integrity, turning every match into a clash of models rather than players. Both miss the point. In a closed league, AI does not democratise and does not destroy. It reinforces the power structure that already exists. I learned this from a small experience years ago. In 2026, while sitting in the commentator's chair at a summer tournament, I mispronounced the name of a legendary player three times in succession during a match. Social media flooded with mocking comments, and I lost my rhythm entirely. After the game, I spent a full month rolling through 48 matches of that team across two seasons, noting every jungle path, every teamfight rhythm, every childhood story of every member. I realised what I lacked was not the skill of reading games. What I lacked was systematic preparation, done consistently, with nothing skipped. The wrong name on screen, the right lesson for a lifetime. And that lesson applies directly here. A big team can hire an entire analysis room of ten people, working around the clock, documenting as meticulously as I once did for a month. A small team does not have ten people. They have two. When iTero sells a tool to GIANTX on exclusive terms, it is not handing the advantage of a ten-person analysis room to a two-person team. It is handing it to a team that already has resources. The gap does not narrow. It widens, at the speed of software. This is the point where I want those chanting about an AI revolution to pause for a second. The revolution does not arrive for everyone at once. It arrives first for those who sign exclusive contracts. Romanticising AI as a neutral force is a mistake with a concrete cost. That cost is not the software licence fee — that sum is small against a professional organisation's budget. The cost lies in cumulative distance: a team with an analytical advantage across three consecutive seasons builds a store of tactical knowledge rivals cannot replicate with money. And in a league with no relegation, no mechanism forces that store to be shared. The question of whether iTero can be copied — the subject of the first section — is therefore not merely a business question. It is a question of whether a structural advantage should be allowed to exist. If the answer is that it can be copied, the problem dissolves over time. If the answer is that it cannot, league organisers will be forced to choose between two paths: mandating equal access for all teams, or restricting the tool — exactly as they progressively regulated coach communication during matches. Both paths lead to the same place: policy. And policy is slow. It is always slower than technology. That is the paradox of this phase — the tool is already here, but the rules of play are not. There is one thing I want to say plainly, and I say it as someone who has sat on both sides of the curtain: inside the war room and in the commentary chair. What you see backstage is not what the public sees on screen. On screen, people see beautiful plays, reverse sweeps, victory embraces. Backstage, people see contracts tightened, three-in-the-morning meetings, decisions made because there was not enough money to choose otherwise. When an organisation signs an exclusivity deal, it does not sign because it believes in the technology. It signs because it believes rivals will not have the same thing. That is the nature of advantage in professional sport. It does not need to be fair. It only needs to be effective. And that is why the question of AI in esports will not be settled by engineers. It will be settled by lawyers. I once watched a team drop their heads onto their keyboards after a loss they were expected to win. It was an October evening in Busan, years ago. I sat in the press row, watching people who had fought to the final minute, and I understood that no technical analysis on earth could explain the feeling of that moment. That night, I wrote a three-thousand-word piece, criticising no one, blaming no one. I only tried to place myself inside their feeling. Busan at four in the morning, a dream shattering into sobs inside a headset. I recount that story because it is why I do not write this piece as an advertisement for technology. What is happening with iTero and GIANTX is not a story about the bright future of artificial intelligence. It is a story about people — about the teams who will win and lose in an arena whose rules are written after the game has already begun. And in that story, people have inadvertently forgotten that players are human, teams are organisations, and wins and losses are only the tip of the iceberg. What I carry away from that interview is not a prediction about the future of AI. It is a question I want league administrators to answer for themselves before the next season begins: if you allow one member of a closed league to own a tool others do not, are you running a competition or legitimising a stratification? Jack Williams and iTero may be the first to place that question on the table. But the ones who must answer are not them. It is those who hold the pen signing league charters — the ones who always arrive late, pen in hand, after the technology has already taken root. Tears do not belong to any brand. They belong to those who believed this arena was fair. And that belief, each time it is sold away in an exclusivity clause, quietly thins by a little.