AI Agents Racing for Restaurant Reservations Spark Debate on Markets and Platforms

AI agents have started grabbing scarce resources like restaurant reservations on users' behalf, sparking concentrated discussion of real-world race conditions, gray markets, and platform change. Real cases show platforms fighting back: booking platform Resy banned the AI booking agent Instinct, while cruder reservation bots have long existed. Several writers expect high-end dining reservations to shift toward auctions, lotteries, and other new allocation mechanisms.

Confirmed

• signulll observed: when large numbers of agents book limited-supply resources (restaurant tables, concert tickets, apartments, flight upgrades) on people's behalf, software-world race conditions enter real life—"two thousand agents refreshing the same restaurant's booking page at once," where the lock is a dinner table for the evening.
• omooretweets noted the restaurant reservation market was already troubled before AI: scalper platforms buying prime-time slots from restaurant staff, merchants paying for higher rankings, and other abuses were rampant; platforms like Resy may be squeezed out if they can't offer new value to consumers or restaurants.
• m5 relayed Casey Collins's report: Resy has long been plagued by bot bookings; the well-known AI booking agent Instinct got banned precisely for bombarding its service with high-frequency requests, while cruder reservation bots have existed for ages.
• Journalist jbahrdestefano (relayed by firasd) laid out several paths for scarce-inventory sites: ban bots entirely; build separate rate limits, authentication, and endpoints for agents; or the platform builds its own sole official agent.
• venturetwins envisioned agent-to-agent deals: an agent holding a coveted reservation could negotiate with a buyer's agent and close for cash or barter; given enough context, LLMs are already smart enough to infer the other side's willingness to pay.
• kolencherry argued AI agents will push popular restaurant experiences toward auctions and lotteries; mattslotnick questioned how this differs from existing sneaker- and watch-bot dynamics, and kolencherry responded that agents generalize the "sneaker bot" model to all scarce experiences.

Why it matters

• The discussion reveals the first real frictions of the agent economy: first-come-first-served allocation of scarce resources breaks down under automated agents, forcing platforms to choose among bans, segregated channels, and building their own agents.
• Resy's ban on Instinct is a landmark case of head-on conflict between a platform and AI agents; meanwhile, existing gray-market practices (staff reselling reservations, paid rankings) show the problem isn't unique to AI, though agents drastically lower the barrier to entry.
• Visions of agent-to-agent trading, auctions, and lotteries foreshadow a potential rewrite of how scarce services are priced and allocated.

2026-09-07 ~ 2026-09-08 · 6 related posts

Full story(2 episodes)→

• Episode 1: AI Agents Racing for Restaurant Reservations Spark Debate on Markets and Platforms(2026-09-07, 6 posts)
• Episode 2: Ten Thousand Vibe Coders Built Resy Reservation Bots(2026-09-08, 2 posts)

• Reservation platforms like Resy were crumbling pre-AI, and agents may finish the job — omooretweets · 2026-09-07
• [source] Debate: AI booking agents will push hot restaurant reservations to auctions and lotteries — matt_slotnick · 2026-09-07
• [source] Agents Fighting Over Finite Supply: Race Conditions Coming to Real Life — signulll · 2026-09-07
• When everyone runs an agent: scarce-inventory sites face five paths, and reservations may have no fix — firasd · 2026-09-08
• Agents bidding against agents: LLMs can already infer willingness to pay, says analyst — venturetwins · 2026-09-08
• [source] Resy long plagued by booking bots; AI agent Instinct banned for hammering its API — matt_slotnick · 2026-09-08