Geo and Jurisdictions: One Brand, Different AI Answers

FoeGlass is an AI-visibility index for iGaming, and one of its core findings is simple: an AI model answers the same question with different brands depending on where the player is. Ask ChatGPT, Gemini, Perplexity, Grok, or Google AI Overviews "where can I play online" and the operators they name change by jurisdiction, because iGaming is regulated per jurisdiction, not per country. A brand can be recommended in one geo and completely absent in the next. FoeGlass measures that visibility at the level where licensing actually decides who gets named.

Why jurisdiction beats country as the unit of measurement

Countries are a convenient label, but they are the wrong unit for iGaming. Licensing is granted, and enforced, below the national line. AI models have learned this: their answers track the legal reality, not the map. So FoeGlass models the world as jurisdictions, not countries.

• USA: New Jersey is a licensed market where AI confidently names DraftKings, FanDuel, and BetMGM. The rest of the United States is a different story, and the same question returns a different, often more cautious answer.
• Canada: Ontario is a regulated market with its own named operators. Canada ex-Ontario is grey, and AI hedges or points elsewhere.

Two prompts, identical wording, different player location, and the brand set flips. If you measure at the country level, you average these apart and see nothing useful.

How one brand can be strong in one geo and invisible in another

Because AI answers follow licensing, a brand's AI visibility is not a single number. It is a grid. An operator licensed in New Jersey may be named as a top pick there while being entirely absent from answers for players in a grey market, where the model either warns, stays vague, or recommends a competitor who does hold a local license.

This is why a global "we rank well in AI" claim is meaningless for iGaming. The honest question is always: named well, for which player, in which jurisdiction, for which intent. FoeGlass answers per cell, not per brand.

How FoeGlass reads a jurisdiction

For every jurisdiction FoeGlass runs the queries a real player would type, in the required language, across the AI engines, and for prepared personas: beginner, regular, high roller. For each named brand it extracts role (recommended, mentioned, or warning), tone, position in the answer, and the sources cited. The data is collected manually by real people, from the required geos, then marked up by two independent AI models, one extracting and one verifying, after the human.

Every cell in the grid has an honest state:

• Present: measured, the brand is named, it has a score.
• Absent: measured, a market exists here, the brand is not named. That is an honest zero, and the deposit goes elsewhere.
• No market: measured, but there is no real market in this jurisdiction, so the cell is excluded from the math.
• Not measured: never measured. This is the absence of data, never shown as a bad result.

A red verdict for a jurisdiction needs a witness: at least one tracked brand named confidently and non-negatively in the same answer. Without a witness, the run is treated as no market and leaves the denominator. That rule keeps a jurisdiction where AI simply refuses everyone from being counted against a single brand.

Why per-jurisdiction measurement is essential now

The stakes are rising because AI answers increasingly replace the click. Google AI Overviews now appear in about 60% of US searches, up from roughly 25% in late 2025. When an AI answer shows, up to 83% of searches end with no click, and up to 93% in Google's AI Mode. If the AI names a licensed competitor in New Jersey and stays silent about you, that player never reaches your site. Averaged nationally, that loss is invisible. Measured per jurisdiction, it is a line you can act on.

The score itself reflects this: it is a weighted average of role, tone, position, and source quality on a 0 to 100 scale, where money intents weigh more than informational ones. Read one jurisdiction at a time, and the pattern of where you win and where you vanish becomes obvious.

If you want to see your brand's AI visibility jurisdiction by jurisdiction, a FoeGlass report maps exactly where you are named, where you are absent, and where the market is being won without you.