AI Coaching in Esports: Where Does the Line Between Advantage and Inequality Sit?

That night I replayed the footage four times. A BO5 at a Southeast Asian event I was watching via stream, where Team A lost the first two games in nearly identical fashion — the same mid-lane push timing at minute seven, the same blind spot in the upper jungle, the same rotation toward the weak side. By game three, Team A rotated completely differently and won narrowly. Game four, they repeated it. Game five, Team B collapsed.

I asked a coach I know. He said an analytics tool had flagged the opponent's blind spot during the break between games two and three. No AI wrote the strategy automatically. It was just a data-suggestion tool. But the question surfaced immediately, and it had nothing to do with whether AI was cheating.

If that tool is exclusive to one team, is that series still a two-sided contest?

That question now has a name: iTero. It has a person answering: Jack Williams. And it has a contract as its backdrop: GIANTX.

Jack Williams is the person behind iTero, an analytics platform supporting esports coaching. He recently gave an interview about how his tool operates, about the exclusive relationship with the organisation GIANTX, and about the likelihood of being copied. The interview has two notable sections. The first discusses working exclusively with GIANTX and the likelihood of being copied by rivals. The second discusses AI-assisted cheating.

Framing it this way is itself a statement. It shapes the story into two poles: commercial exclusivity at one end, competitive integrity at the other. In between sits a gap both sections skip — league fairness.

GIANTX, as far as I know, is an organisation present in the League of Legends EMEA ecosystem, formed through the merger of two older organisations. It is a permanent member team in a franchised league — no relegation, no elimination by qualification. That characteristic matters more than people think, and I will return to it.

On the tool side, what is iTero? No data on sample size, accuracy, processing speed, or evaluation method is disclosed in the piece. That was the first thing I noticed. A tool promising to change match preparation, with no number attached. I do not dismiss it for lacking numbers. I log the absence of numbers as a fact. The mistake in Surabaya taught me to question data, not trust it.

Now to what I consider the most important part, and the part both the interview and most commentary on it are reading in the wrong place.

The real story here is a commercial variable the esports industry has no governance framework to handle: match-preparation tools are becoming exclusive assets, while the existing rulebook was written only to govern in-match behaviour.

To understand why, separate 'AI in esports' into three moments that differ by regulation.

The first moment is in-match. Real-time assistance — software reading match state and offering suggestions during play — is explicitly banned in every major title. No grey zone. Nothing to debate. If the debate sits here, it ended years ago.

The second moment is post-match. Retrospective analysis — reviewing heatmaps, ward positions, item timings — is widely accepted. Every professional team does this to some degree, some with people, some with tools.

The third moment, and this is where the real battle happens, is between games in a BO3 or BO5. This is a short window, often ten to thirty minutes, when a team must decide whether to change strategy based on what just happened. What happens in this window is not governed by the in-match assistance ban, because the match has stopped. But it is not quite pure post-match analysis either, because the result still has immediate value.

This is the true grey zone. And a tool that can shorten the time from previous-game data to next-game adjustment has value that does not lie in knowledge — it lies in speed.

That is why I say the interview reads the wrong place. If iTero is a tool helping a team read the game faster within the between-game window, its advantage is not 'knowing more than the opponent'. Its advantage is knowing faster within a timeframe both teams are bounded by equally. That timeframe is a scarce resource. And whoever controls how time is compressed controls part of the game.

Now the title variable. This is where technical analysis must separate out, and I have not seen anyone do it.

Major titles do not run on the same update cadence. Some titles run slow and large — systemic changes every few months, with long stability stretches in between. Some run fast — small tweaks every two weeks, continuously shifting the meta.

For slow-changing titles, the value of an AI coaching model is depth: it can learn from long history and still be right. For fast-changing titles, that value inverts: historical depth depreciates quickly, and what remains is the ability to detect a shifting meta earlier than rivals.

Two different value propositions. Two fundamentally different products, even packaged under one name.

If a platform markets the same solution for both types of titles, that is a signal to ask questions. Not a sign of cheating, but a sign of a marketing claim exceeding what it can prove. I have seen this in football for years: the same expected-goals model sold to a counter-attacking side and a possession side, as if context did not exist.

And here is where I want to be blunt about the 'exclusivity' section of the interview.

An exclusivity deal is not a cheating issue. It is a structural issue. And league structure determines its severity.

In an open-circuit league — with qualifiers, relegation, and season-to-season slot changes — structural advantage tends to erode. Weaker teams get eliminated, stronger teams rise, and any advantage that cannot convert into results is filtered out over time.

In a franchised league — fixed members, no relegation, guaranteed slots — structural advantage does not erode. It persists season after season. A member team with exclusive access to an analytics tool, over four consecutive years, accumulates a data gap no one claws back through pure competitive effort.

GIANTX, if indeed a member of the EMEA franchised model, has its exclusivity deal with iTero sitting precisely inside this structure. And that is why the gap between the interview's two sections matters: it is the gap named 'league fairness'.

The question neither interviewer nor interviewee asked is: if a match-preparation tool can change outcomes, what responsibility does the league operator bear?

History has answered a similar question many times. When in-match coach-team communication became effective enough to raise fairness doubts, leagues progressively limited it. When voice comms became an advantage, they limited time. When screen-sharing became a problem, they changed the rules. In every case, the principle was the same: when a tool changes outcomes, the operator must choose between mandating equal access or restricting the tool.

With AI coaching tools, both directions have problems. Mandating equal access means the operator must buy or license the tool for every team, and must define 'AI coaching tool' clearly enough not to inadvertently ban harmless things. Restricting the tool means drawing a line between 'permitted analysis' and 'banned assistance' during a period when the technical line shifts every month.

Both are hard. But ignoring it is not defaulting to neutrality. Ignoring it is defaulting to allowing the exclusivity deal to continue.

Meanwhile, the 'likelihood of being copied' section reads like a commercial worry. It is indeed a commercial worry. But placed beside the integrity section, it accidentally reveals something more interesting: toolmakers understand their advantage is temporary. They know rivals will copy. They know the market will saturate. And the only question left is who regains the advantage faster — the monopolist wants to hold it as long as possible, everyone else wants it gone as soon as possible.

World Cup 2026 lifted the trophy with tackles no one remembers. In esports, the equivalent 'tackles no one remembers' are ten seconds of earlier rotation, of jungle path changes based on a behavioural pattern no one noticed in the stat sheet. The AI tool does not create those moments. It just finds them faster.

And here I must address what the interview avoids by placing two topics side by side.

The common reading of this story is a worry about cheating. Will AI write strategies for coaches? Will it read the match in real time?

I think that worry reads the threat wrong. It focuses on a behaviour already explicitly prohibited, while the real danger lies in a behaviour fully permitted.

In-match cheating is a detectable, punishable, provable offence. An exclusivity deal between a platform and a member team is a legal, public contract violating no existing clause. It is not an integrity issue. It is an opportunity issue.

I am not downplaying integrity. I am only saying the harder problem is being skipped because it has no clear villain. No one cheats by signing a contract. No one gets banned for building a product.

And here I want to pose the reverse question to those worried about AI cheating: if you fear a tool that can change match outcomes, why do you not fear that tool existing only in one team's hands? Detected cheating gets handled. Structural inequality does not get handled because no one has the authority to handle it.

There is one more detail in the interview I read as a signal, not as content. It is the toolmaker choosing to worry about being copied rather than worrying about the league operator redefining the legal boundary of the tool. He believes his advantage is technical. But the real advantage of an exclusive relationship is the contract — and a contract can be overridden by league rules at any time.

What I will track in the next cycle is not what new feature iTero has. I will track how franchised leagues respond to this kind of exclusivity deal — whether they mandate equal access, restrict the tool, or stay silent.

Because when a fixed member team monopolises a preparation tool, the question is no longer which team prepares better. The question is whether that league is still a contest between teams, or has become a contest between contracts. And if the answer is the latter, every standings table we read needs a new footnote.