Jack Williams, iTero, and the Undrawn Line of AI Coaching in Esports

Jack Williams, iTero, and the Undrawn Line of AI Coaching in Esports

I still remember the feeling of sitting in front of the screen, replaying footage from an LEC summer BO5. Between game three and game four, the camera cut to the GIANTX team room. Nothing was loud. There was no "quadra kill" shout, no applause. Just a coach leaning over a laptop, hands moving across a data sheet I could not read from a distance, and in those seven short minutes perhaps a machine-learning model was suggesting something — about the enemy jungler's clear speed, about the win rate of a mid-lane duo, or about the probability that a Baron play would succeed. The arena stayed silent. I sat up straight as if someone had pressed a revive button, because I realised I was watching something that still had no formal name in the rulebook: a coach consulting a machine.

That is when the story of Jack Williams, iTero and GIANTX began to haunt me. Not because the technology is new — esports has lived with data analysis for a decade. But because the line between advantage and cheating is being stretched thinner every day, and Jack Williams is the first person willing to sit down and face that question under his own name.

Context: a name from the backstage

Jack Williams is not a household name. He comes from behind the curtain — the world of people who build tools, sell solutions, and sit at the intersection of technology and esports. iTero, the product he is tied to, belongs to the AI-coaching category: systems that collect match data, build models, and feed suggestions to coaching staff for pre-match drafting, between-game adjustments, and post-match analysis.

The interview I followed revolves around two axes. The first is commercial: iTero works exclusively with GIANTX — an EMEA organisation widely reported to compete within the LEC ecosystem, born from a historic merger of two gaming organisations — and Jack Williams acknowledges the likelihood of being copied. The second is ethical: whether AI tooling can become a vehicle for cheating.

These two axes look separate but are two faces of one coin. Selling an exclusive tool to one team means deliberately creating a competitive advantage. When that advantage grows large enough, the question of where the limit lies surfaces on its own — not from a court, but from the community. And the esports community, as I have learned over the years, does not forgive what it cannot verify.

Based on my experience following matches from the early LCK years through recent international events, I see a clear pattern: whenever a new tool or rule appears, the community does not react to the tool first. It reacts to the question of who gets access to it. That is why the iTero and GIANTX affair cannot be read as a pure technology story. It is a story about power.

Where the real value of a coaching machine lies

To understand why an AI coaching tool can be so controversial, its value must be split into two different capabilities — with opposite durability.

The first is speed of patch-solving. League of Legends runs on a two-week update cadence. Each patch shifts balance, changes champion strength, and rapidly devalues patterns learned from the past. In that environment, AI's value is not knowing the answer but detecting the meta shift a few days before opponents. That is a tempo advantage, not a knowledge advantage.

The second is depth of historical modelling. Dota 2 runs on a very different rhythm. Valve patches less often, but each change is more systemic and more destructive. Long stretches of stability sit between major changes, letting models trained on historical data retain value longer. Here AI can build a deep picture of one patch instead of racing against it.

This contrast leads to an inference the interview itself does not state: if iTero markets an identical product across both titles, that is a red flag. A tool optimised for speed cannot simultaneously be optimised for depth. Its value must invert according to the patch cadence of the title it serves.

This matters because Jack Williams is selling a universal promise. And universal promises, throughout esports history, hit the same wall: non-uniform data. A model trained on tens of thousands of ranked games cannot say much about a BO5 final where two teams spent two weeks preparing specifically for each other. This is the foundational paradox of every analytical tool: the more abundant the data, the less it relates to the specific match in front of your eyes.

The real issue is the word "exclusive"

The key to this story lies in the exclusivity contract, not the algorithm. GIANTX is a member of a closed, franchised league. In a closed league there is no relegation, every member is a permanent member, and any structural advantage one member holds persists across seasons instead of being competed away.

If an analytics tool genuinely affects match outcomes, exclusive access to it becomes a protected form of preparation inequality. In an open system, that advantage would push other teams to buy similar tools and close the gap. In a closed league, there is no natural equalising mechanism. The league operator — Riot Games, in this case — is the only party able to intervene, either by mandating equal access or by restricting the tool.

I have seen this script before. It is what happened as in-game coach communication was gradually tightened over the years, from allowing coaches behind players, to a few seconds of talk during breaks, to stricter limits still. Each tightening split the community in two: one camp arguing the league was strangling strategy, the other arguing it was protecting integrity. The battle over iTero and AI coaching is the next chapter of that book — except this time the coach's counterpart is a machine-learning model rather than a human who knows fear.

And Kazan taught me that neither the ball nor the Baron forgives those who tremble. A machine forgives even less. It does not tremble, does not know an empty arena, does not know the pressure of a Baron call that decides an entire season. That is why it can be useful. That is also why it is frightening.

The real grey zone: seven minutes between games

Public debate about "AI cheating in esports" usually targets the image of a player being whispered to mid-match. But real-time in-game assistance has been unambiguously banned in every major title for years — to the point there is nothing left to debate. The interesting and dangerous grey zone is the window between games in a BO3 or BO5.

In those seven minutes, the line between "analysis" and "coaching" blurs. A human coach in seven minutes can recall a few things, sketch on a board, persuade the team to adjust top lane. An AI tool in those same seven minutes can produce a personalised list of suggestions for each player, based on thousands of modelled similar situations. The difference is not the nature of the act. The difference is scale and speed.

This is the largest blind spot in the whole debate: current rules were written for a world where the human is the slowest information processor in the room.

When a machine processes hundreds of times faster than a human yet still sits inside a legally permitted break window, the question is no longer whether the tool is legal. The question becomes whether the notion of fairness we are using still fits. That is a question no league operator wants to answer publicly, because any answer will anger someone.

I once watched Damwon Kia destroy DRX 3-0 in the 2026 LCK Summer final, when Canyon and Beryl controlled seventy-five percent of map vision in the first fifteen minutes of each game. Back then I measured their heart rates and reaction times from replay clips and found that playing to an empty arena helped them focus. Damwon's silence was not emptiness; it was the waiting room of history. The iTero story is a variation on the same theme, except this time the silence does not come from the stands — it comes from a model that never speaks and never explains why.

A machine cannot be argued with, and that is the real bad news

Now I want to invert the familiar argument. People fear AI coaching because it may exceed human physical and mental limits. But that fear places the focus in the wrong place.

What makes a tool like iTero a problem is not that it is too strong, but that it cannot be challenged. A human coach who makes a wrong call can be questioned, confronted, replaced. A machine-learning model that makes a wrong call simply stays silent. You cannot ask a model why it advised two bruisers in mid lane and receive a verifiable answer. You can only accept the result. This is something modern esports teams do not yet have the cultural tools to handle.

In other words, the biggest risk of AI coaching is not cheating. It is the migration of responsibility. When a strategic decision originates in a black box, no one is accountable when the team loses. The coach says the model suggested it. The analyst says the data indicated it. And the player — the one who actually walks onstage to take the hits — becomes the only person who cannot blame anyone else.

The copying risk Jack Williams admits to is not merely a copyright story. It is a cultural story: if iTero is exclusive to GIANTX, hundreds of other teams will try to recreate that tool by hand, in Excel, through analysts working eighteen-hour days. That race produces a generation of players and coaches judged not by competitive instinct but by dependence on tooling. And I find a generation struggling there, exactly as I once found one in the analyst columns grading top-lane matchups through an LCK summer.

What no one has written, and what should be written

There is an angle both disclosed interview headings skip: the league's angle. The first heading covers exclusive work with GIANTX and the likelihood of being copied. The second covers AI-assisted cheating. Between them lies an unnamed gap: whether a franchised league should let one member hold a tool the others do not.

When a league runs as a closed club, every resource asymmetry becomes a permanent asymmetry. A team can buy the best coach, the best facilities and, someday, the best AI model. The difference among these three lies only in how hard they are to measure, not in their nature. If we accept that a team may pay for a better coach, there is no principled reason to forbid paying for a better model — unless we admit that what is being protected is not fairness but a romantic idea of fairness.

And here I must be careful with myself, because I know my habit of turning every dispute into an elegy. The truth is less poetic. If a tool truly creates a measurable win-rate gap, teams will not wait for the rules to change. They will start building their own versions, hiring people who can build models, and turning analytics rooms into engineering offices. That process has already begun, and Jack Williams's interview simply says out loud what most coaching staffs whisper to each other in arena corridors.

What to remember

If you are a team without an exclusive iTero deal, the first thing to do is not to copy the tool. The first thing is to define clearly what kind of edge you need: speed of patch-solving, or depth of preparation for a specific opponent. These two edges require two different tools, and no single model delivers both at once.

If you are a league operator, the worthwhile move is not to ban AI coaching — that is unenforceable and counterproductive in PR terms. The worthwhile move is to define which time windows permit tool use, who can access data, and which data must be published to every team. Transparency about time windows will resolve most disputes, because most disputes are not about algorithms but about nobody knowing what happens in the seven minutes between games.

Damwon 2026 showed that the greatest glory can sprout from empty arenas. Perhaps AI coaching shows something similar in reverse: our greatest fear can sprout from a silent room where a machine quietly rewrites the rules of the game before anyone has read them. And if you watch a BO5 next time, pay attention to the break between game three and game four. Those seven minutes now contain far more than a glass of water and a few words of encouragement. They contain the future of this sport, being decided in silence, one suggestion at a time.