AI Coaching at the Commercial Boundary: Jack Williams, iTero and the Exclusive GIANTX Deal

GIANTX holds something the rest of League of Legends EMEA does not: exclusive access to an esports coaching tool built on artificial intelligence. Jack Williams, the man behind iTero, has just given an interview about that very deal — the exclusive partnership with GIANTX, the likelihood of being copied, and a larger question sitting behind it: is artificial intelligence cheating in esports?

In Seoul, I reopened an old notes file of mine. In 2026, when stadiums around the world sat empty, I counted the online viewership of K League 1 matches one by one and arrived at a simple conclusion: the value was never in the stands, it was in the stream of data moving around them. I started listening to data when the stands went quiet, and it told a completely different story. Williams's interview tells that same story, just in a different sport.

Context: from whiteboards to machine learning models

Esports coaching has passed through three generations of tools. The first was the whiteboard and the recording, where a coach rewound VODs and drew movement paths by hand. The second was the match-data platform, where every metric from stage-by-stage win rate to ward placement was digitised into spreadsheets. The third, the one iTero represents, puts machine learning models into the preparation itself: predicting pick-ban trends, finding patterns in opponent behaviour, suggesting adjustments between games.

The regulatory line is not new. In League of Legends, Riot Games removed coaches from in-game communications years ago; all assistance is permitted only in the windows between games and before a match. Valve's Dota 2 has a different tradition regarding third-party intervention. That means the grey zone worth debating is not real-time assistance, which is clearly banned, but the between-game window, where a tool can change the outcome of a BO3 or BO5 series.

GIANTX, according to industry reporting, is an EMEA organisation formed from the merger of Excel Esports and Giants Gaming, competing within the LEC system. That is a closed league: no relegation slots, fixed members by season. The question of tool exclusivity therefore carries very different weight than it would in an open system, where a weaker team can be relegated and structural advantages dissolve on their own.

One thing needs to be said plainly about the material. This interview is a B2B thought-leadership piece, not a tournament report. It contains no bracket, no format, and not a single performance number. All it provides is two section headings — the exclusive GIANTX partnership and the likelihood of being copied — alongside the theme of AI-assisted cheating. That is enough to reconstruct the industry logic, but not enough to draw conclusions about the product.

The image of Natus Vincere lifting the Aegis of Champions at Gamescom fourteen years ago appears in the article as a nostalgic detail. It carries no analytical meaning about current patches or formats. Read it as a signal about the contemporary Dota 2 landscape and you have misread the layer.

What is actually being sold here

The first point to separate out: the value of an AI coaching tool inverts completely depending on the patch cadence of each title. League of Legends patches every two weeks. The lifespan of any data pattern is short, and AI's edge shifts from "solving the meta" to "detecting the meta delta faster than opponents". That is a speed advantage, not a knowledge advantage. Dota 2 runs on a different rhythm: major updates are less frequent but restructure the game harder, interspersed with long stable stretches. There, models trained on historical data retain value longer, and depth of modelling beats speed.

A single product marketed identically across both titles is a red flag, because its competitive advantage must be defined in opposite ways in the two patch environments.

The second point concerns the structure of exclusivity. In a closed league, a member's structural advantage is not competed away across seasons. No relegation means no natural mechanism erases the gap. An exclusive analytics-tool agreement therefore does not resemble a sponsorship contract. It sits in the same regulatory category as any other match-preparation advantage, and its consequences accumulate by season rather than dissolving match by match.

The third point, and the one Williams touches when he speaks about the risk of being copied: iTero's moat is not in the technology, it is in the contract terms. A machine learning model, once proven effective, can be replicated within months by any sufficiently capable engineering team. What cannot be replicated quickly is private access, and that is precisely what GIANTX is paying for.

Tool exclusivity is a commercial advantage with an expiry date, and it expires faster than the speed of technological copying.

The fourth point sits on the publisher's side. When a tool becomes powerful enough to affect competitive outcomes, the league operator faces two choices: mandate equal access for all teams, or restrict the tool. The history of Riot Games regulating in-game communication shows this precedent is familiar. Both choices erode the value of the exclusive deal iTero is selling.

On the data side, any assessment of iTero's durable edge is missing a critical piece: the specific patch cadence, the tournament server lock rules, and the data-availability windows. Without those three parameters, every performance claim is just a claim. In the material I reviewed, not one sample figure, evaluation methodology, or dataset size was disclosed.

Contrarian angle: the wrong framing

Most public debate circles around the word "cheating". That is the wrong frame, or at least a narrow one. Real-time assistance is banned in every major title, so the interesting part is not there. The interesting part is the middle ground: a legal, licensed tool that only one team has.

The right frame is resource fairness inside a closed league. When a fixed member holds a tool exclusively, the league has implicitly chosen to permit inequality in preparation. This is the least examined angle, because it sits between two more marketable headlines: commercialisation and cheating.

One more counterpoint deserves mention. The narrative of "AI replacing coaches" is compelling but off-centre. AI does not replace the person reading the game. It replaces the speed of reading the game. And in an environment patching every two weeks, speed is the entire game. Data gives me the map, but instinct is what picks the road.

What remains after the interview

If a tool is strong enough to decide a BO5 series, the league operator will have to act. The question is not whether that happens, but who will be first to publish the metric proving it: the publisher, the team holding the exclusive rights, or an independent analytics group on the outside. A tool's value is not priced on the server, but in the operating system around it. A contract is only truly complete when its story is told the right way.