Jack Williams, iTero and GIANTX: When AI Enters the Strategy Room, the Boundary of the Rules Begins to Blur

One late October evening, as I sat in the editing booth after a group-stage livestream, my phone buzzed. A friend who does data analysis for an EMEA organisation wrote: "Have you seen iTero? GIANTX just signed it exclusively." I read the message, then read it again. Not because the name iTero was unfamiliar — but because of the word "exclusively". In nine years observing this industry, from tournament organiser to large-event host, I have watched teams fight over coaches, players, and sponsorship deals. But fighting over an artificial-intelligence analytics tool, and signing it exclusively — that is a new kind of competition. And like every new kind of competition, it touches a boundary the rules have not yet drawn.

Jack Williams, the figure behind iTero, has just held a public conversation on exactly this subject: working with GIANTX exclusively, and the likelihood of being copied. Alongside it sits another section — AI-assisted cheating. Reading those two headings side by side, I realised they are not two separate stories. They are two ends of the same thread. At one end is a commercial question: how do you protect the competitive edge of a product whose very nature is data that can be imitated. At the other end is an ethical question: at what boundary does an assistance tool become a cheating tool. And in the middle — the space nobody wants to discuss — sits the question of fairness inside a closed league.

Context: When did AI slip into the strategy room

To understand why the iTero story matters, it must be placed in the context of the annual season. Esports is no longer a playground of pure reflex. At the top tier, the gap between teams in individual skill has been compressed to almost nothing. When every player in a top league can pull off an elite play, the deciding factor shifts to preparation: reading the patch, building draft scripts, managing stamina, and adjusting tactics between games in a BO3 or BO5. That is precisely the gap through which AI analytics tools slip in.

Technically, products like iTero sit in the middle zone: they do not interfere with a live match in real time, but process data pre-match, between games, and post-match. Pre-match, they aggregate head-to-head history, pick-ban trends, and tactical patterns. Between games, they suggest adjustments based on what has happened. Post-match, they turn piles of VOD and statistics into condensed conclusions so the analytics staff need not rewatch every second. It sounds harmless. And that is exactly why it is dangerous from a governance standpoint: it sits right in the grey zone the rules have not clearly defined.

Here lies a difference between titles that I consider a first-order commercial variable, yet it is almost never mentioned in discussions. Dota 2 operates on a cadence of large, infrequent, disruptive patches — huge systemic updates followed by long stretches of stability. League of Legends is the opposite: a patch every two weeks, rotating continuously. For Dota 2, an AI model trained on historical data retains value over a longer window — the edge belongs to depth of modelling. For League of Legends, the lifespan of any pattern is short, so AI's value shifts from "solving the meta" to "detecting meta drift faster than opponents" — a tempo advantage, not a knowledge advantage. The same product, sold with the same pitch across both titles, is a red flag. If iTero does not tier its product by each title's patch cadence, then either it is selling something generic, or it is hiding something about the tool's real limits.

GIANTX, from what I know of the EMEA picture, is an organisation with a foothold in the European League of Legends ecosystem — where Riot Games sets the rules. That places the exclusive arrangement between GIANTX and iTero within a specific legal framework: Riot's third-party software and competitive-integrity rules. This is not a small detail. It determines whether the arrangement is valid at all, and if valid, how far that validity extends.

A note on sourcing is also needed. The conversation with Jack Williams is a B2B thought-leadership piece for industry insiders, not mainstream news. It provides no performance data, no sample size, no evaluation methodology. That means every claim about how much iTero actually helps a team win more is unverifiable from this source. I must state that clearly before analysing, because bold experimental discipline does not mean fabricating evidence.

Core analysis: Exclusivity, copying, and the commercial trap

The value of an AI coaching tool is not in the algorithm, but in exclusive data and exclusive speed — and both are extremely easy to copy.

Let us split the problem into three layers. The first is the model: the machine-learning algorithm processing data. This layer can be copied but takes time and good engineers. The second is data: the VOD archive, scrim history, head-to-head records. This layer is the real boundary — whoever owns proprietary data wins, because a model trained on public data can be built by anyone. The third is workflow integration: how the tool meshes with a coaching staff's habits. This is the layer that creates durable advantage, and also the hardest to copy, because it lives in people's heads rather than in code.

When GIANTX signs exclusively with iTero, what they are really buying is not the algorithm — it is early access and the right to shape the product around their own data and process. Exclusivity in esports is unlike exclusivity in manufacturing; it is exclusivity over time, not over goods. Competitors can still buy a similar tool six months later, but six months in a season is the distance between top four and the knockout stage.

But this is where copyability becomes a real threat. Because the second value layer — data — is largely public in esports. Official match VOD is viewable by anyone. Draft data is collectable by anyone. The only thing not public is internal scrim data, and that is precisely the asset an exclusive deal like GIANTX–iTero tries to lock up. If a competitor buys a rival tool and feeds it their own scrim data, the model gap can narrow quickly. The question Jack Williams raises — the likelihood of being copied — is really a question about whether the data advantage runs deep enough to withstand rivals building their own. And the answer, based on my experience watching matches over many years, is: no data advantage is permanent in an industry where everything is streamed, recorded, and re-analysed by the community.

This is where the "economics" of esports shows itself more clearly than any league table. A tool serving only one team is a tool without economies of scale. AI model development cost is largely fixed — paid once, used across many customers. If iTero only sells to GIANTX, it must price extremely high for a single customer, or it must have an expansion plan immediately after the exclusivity phase. Both possibilities lead to the same conclusion: the exclusive arrangement almost certainly has a term limit. That does not make the story less interesting — it just changes the question from "is exclusivity fair" to "how long does the exclusive window last, and who buys next".

One more point struck me when reading about the two sections mentioned in the conversation. The first — exclusive work and the likelihood of being copied — is about protecting an edge. The second — AI-assisted cheating — is about ethical limits. These two sections look opposed, but in fact they point the same way: both are consequences of a tool running faster than the law. When the law has not clearly defined the boundary, the same technology can be sold as a legitimate advantage to one team, and treated as cheating if another team uses it differently. That ambiguity is not accidental — it is the ideal environment for a vendor of an exclusive tool.

Contrarian angle: The forgotten frame — fairness in a closed league

What bothers me most about this conversation is how it is framed. The two sections place the issue in two frames: commercial and integrity. But there is a third frame sitting between them that almost nobody mentions: competitive fairness.

Think about the structure of a closed league like the LEC. It is a model with fixed franchise slots, no relegation in the traditional sense. In such a system, structural advantage does not get "competed away" as in open circuits. If one member has exclusive access to an analytics tool with real influence on results, that advantage persists across seasons, accumulates, and becomes a form of structural inequality. In an open circuit, a weak team can be relegated and return with a new tool. In a closed league, there is no such self-correcting mechanism.

And this is the point I want to push further: a league that permits exclusive tooling is, by nature, choosing to permit preparation inequality. That is a governance decision, even if it is made passively, even if nobody realises they are deciding. Publishers and league operators will soon face pressure: either mandate equal access for all teams, or restrict the tool. History has shown this. Years ago, in-game coach communication was progressively tightened, because people realised that an information channel affecting results must eventually be managed. AI coaching tools are walking that same path, just a few steps slower and in a grey zone legitimised by contract.

But wait — before concluding hastily, I must challenge myself. There is a strong countervailing fact: no evidence shows AI coaching tools create a result gap large enough to be called unfair. There is no data, no sample size, no published evaluation method. Every claim about iTero's effectiveness in my source is unverifiable. So if I say "GIANTX has an unfair advantage", I am going beyond the data. What I can say with certainty is only this: the structure of the exclusive arrangement creates a risk that needs governance, not a proven fact of injustice. This is the line between verified contrarianism and baseless contrarianism, and I must stand on the side of verification.

Then the second countervailing fact, and it is stronger still: exclusivity itself can backfire commercially. If iTero serves only GIANTX, it limits itself to one customer and loses economies of scale. This suggests the exclusive arrangement almost certainly has a term — a time-limited priority contract, not permanent exclusivity. And if so, the "structural advantage" I worried about may be only temporary, a head start rather than a permanent gap. This reverse assumption is mandatory before concluding, and it softens — but does not erase — my initial concern.

One more point deserves mention regarding cross-title transferability. If Dota 2 punishes machine-learning on historical data differently from League of Legends, then a tool that performs well in one title may fail in another without anyone anticipating it. This is a risk an exclusive deal cannot protect the customer from. GIANTX may be buying a window of advantage in the current patch environment, not a universal weapon. And if the patch cadence changes — say Riot inverts its cycle, or adds a mechanism that devalues old models — the advantage evaporates without any rival doing anything.

Finally, I must touch on the "golden moment" I still hunt in this profession. The most newsworthy moments are not when champions lift the trophy, but when a new variable appears and nobody has named it yet. The exclusive arrangement between GIANTX and iTero is exactly such a moment. It has not caused a scandal, no team has filed a complaint, no publisher has spoken out. But it is a seed. And in my experience, seeds like this only sprout when it is already too late to do anything.

Progressive conclusion: The map is only right until the ball touches down

Back to the metaphor I have carried since my early blogging days in Incheon. The map is only right until the ball touches down. An AI coaching tool is a beautiful map — it draws patterns, predicts drift, suggests moves. But the ball still has to touch down. The player still has to press the button. The coach still has to decide in the thirty seconds between two games. And in that moment, no algorithm can replace a human.

Every arena has a map; the winner is whoever reads the map before the ball rolls. But the question the iTero and GIANTX story leaves us is not "who reads the map better". The question is: who is allowed to own the map, and whether owning it is inadvertently reshaping the rules of an entire league. The pitch and the map are not opposed; they are just two ways of drawing the same trap — and the biggest trap right now is that we are letting technology outrun the rules.

I do not think AI coaching is a threat. I think it is inevitable, much as data analytics changed football over the past decade, when clubs began hiring data scientists and turning every pass into a variable. But I want to see a transparent governance framework before it becomes the norm: clear rules on which tools are allowed, which data may be shared, and whether access must be equal among members of the same league. If not, we will have a generation of champions nobody can explain — winners whose title was woven from talent, or from a software contract.

And that is the truly frightening question — not whether AI is smarter than humans, but whether we have the courage to set the rules before the tool sets them for us.