When we talk about competition, one idea seems obvious: competitors are the companies fighting over the same customers. But does that logic still hold when consumers ask an AI for advice? LirenPrism's measurements suggest a new phenomenon is emerging: AIs build their own competitive arena.

A competition no longer defined solely by the market

When a user asks an AI to recommend a product, it does not present the whole market. It selects a limited number of players, who become, for the user, the ground for comparison. In other words, before the consumer even makes a choice, a first selection has already been made.

So it is no longer only companies that define their competitors: AI engines take part in shaping this competitive field too.

What our measurement shows

The mAIr Arena report analyzes 120 answers produced by several AI engines, with and without web search. In this study:

  • the product studied appears in 100% of the answers;
  • the AIs associate it with a recurring set of competitors;
  • 98 different players are mentioned in total;
  • no secondary competitor is systematically cited;
  • the report explicitly states that this is the competitive field constructed by the AIs, not a real ranking of the market.

This distinction is essential. The report does not try to determine which products are best: it measures which products the AIs choose to place in competition with one another.

AIs do not only choose the competitors

AI engines also select the criteria for comparison. In this study, the most frequently used criteria include:

  • composition;
  • performance;
  • versatility;
  • practicality.

Here again, the report reminds us that these criteria correspond to what the AIs emphasize in their answers, not necessarily to what matters most to consumers. In other words, the AI does not only answer a question: it also shapes the way the comparison will be made.

The same company can see its competitive environment change

The measurements also show that AI providers do not all emphasize the same criteria. Depending on the engine used, the differences concern in particular price, practicality, quality, health, safety or support.

The product remains present; what changes is the way it is told. This means that the same product can be perceived differently depending on the AI consulted.

A new form of visibility

For years, visibility mainly meant appearing in the results of a search engine. With generative AI, a new challenge emerges: being visible is no longer enough — you also have to be present in the group of competitors the AI decides to show. Because a product absent from that selection risks never entering the user's comparison stage.

What this finding does not demonstrate

The report does not demonstrate that AIs identify the best products. Nor does it demonstrate that the competitors displayed match the economic reality of the market. It measures only the players and the criteria that AI engines mobilize when answering the same question, at a given moment.

Why this finding matters

The stakes go well beyond the product analyzed. If AI engines become a major gateway to information and recommendation, they also help define:

  • which players exist in the user's eyes;
  • which competitors deserve to be compared;
  • which criteria are used to tell them apart.

The competitive field would then no longer be shaped only by the market, companies or consumers: it would also begin to be shaped by artificial intelligence engines. This is precisely what LirenPrism seeks to measure — not who is truly the leader, but how AIs redraw competition as they present it to their users.

Test for free: create a LirenPrism account, enter code LIREN2026 and get 2 free mAIr.

Test my visibility