DataDome launches Priority Protect virtual waiting room for AI

DataDome rolled out Priority Protect on May 20, 2026, a virtual waiting‑room that separates human buyers, authorized AI shopping agents and malicious bots before they reach the checkout line.

The product builds on DataDome’s existing bot‑detection engine, classifying every request in real time and applying distinct access policies for each traffic class. In a recent midnight ticket sale for a major sporting event, bots accounted for 31% of queue traffic—2.4 million of 7.8 million requests—highlighting the limits of legacy queues that make a single trust decision at the door [2]. Unlike traditional virtual queues that only absorb human‑driven spikes, Priority Protect continuously re‑validates visitors throughout a session, re‑challenging or ejecting them if behavior shifts. The system draws on roughly 5 trillion daily signals from DataDome’s network to power its “intent‑aware” analysis and lets operators set capacity, release rates and custom policies via a dashboard or API [2].

The launch reflects a broader shift as retailers confront “agentic commerce,” where AI‑powered agents can add items to carts, compare prices and lock inventory in milliseconds. DataDome argues that conventional fraud tools intervene too late—after inventory is already reserved—while its new platform aims to prioritize traffic at the moment of demand, ensuring that every spot in line goes to a real customer or a trusted AI agent [1]. The company’s move follows its April 2025 release of large‑language‑model detection and intent‑based AI models, signaling a strategic pivot from treating all automation as hostile to distinguishing legitimate agents from unauthorized bots [2].

Customers such as Etsy, PayPal and SoundCloud have already adopted DataDome’s solutions, and the Priority Protect waiting room runs on the merchant’s own infrastructure with branded templates and estimated wait times, avoiding redirects to third‑party domains [2]. Pricing has not been disclosed, but the product is available now.

If AI agents continue to operate 24/7 and execute purchases at machine speed, every sale could effectively become a flash sale, forcing merchants to redesign queue logic around a three‑way split rather than the traditional human‑versus‑bot binary. Whether retailers can balance fairness, revenue and security in this new landscape remains the key question.