Invisible to the Machine: Auditing AI Restaurant, Cafe, and Bar Recommendation Against a Complete Market Census

Research from arXiv highlights a significant gap in AI-driven local discovery, particularly in the food and drink sector. A comprehensive audit of AI recommendations for restaurants, cafes, and bars in two markets revealed that a substantial majority (85.6%) of existing venues were never recommended by the four leading AI systems tested. This indicates a potential bias and incompleteness in current AI recommendation algorithms, impacting market visibility and revenue distribution for local businesses.

Why it matters

The widespread adoption of AI for local discovery means that a lack of visibility for a large proportion of businesses directly impacts their potential revenue and market share. Organisations relying on or developing AI recommendation systems must address these biases to ensure fair market representation and robust user experiences, while businesses must consider strategies to gain AI visibility.

Key insights

• AI assistants are increasingly the primary interface for local discovery.
• There is a lack of understanding regarding which venues AI systems recommend.
• A census-denominated audit was conducted on 4,776 venues across two markets (Canggu and Ubud, Bali).
• Four production AI systems (ChatGPT, Claude, Gemini, Perplexity) were evaluated using 2,208 search responses to 96 queries.
• A significant 85.6% of venues were never recommended by the AI systems during the audit period.
• This audit method allows for measurement of what sampled audits cannot, due to observing the full market.

Source

arXiv — Computers and Society — https://arxiv.org/abs/2608.07069