3 hours ago
16
McDonald’s has a new name in its drive-thru lane, and it answers to Archy. The fast-food giant has begun testing McDonald’s AI drive-thru ArchIQ, an AI-powered system designed to handle orders and support broader restaurant operations, and early numbers suggest it is already moving at scale.
According to a franchisee account on X, ArchIQ has processed more than one million transactions since testing began. Even more notably, about 90% of those orders were completed without any human intervention. For McDonald’s, that is a clear sign of how far drive-thru automation has advanced inside one of the world’s largest restaurant chains.
The system is currently live at five U.S. McDonald’s locations. That is still a limited rollout, however, it gives the company a controlled test bed before any wider deployment.
McDonald’s AI drive-thru ArchIQ is handling orders and restaurant tasks
ArchIQ was officially unveiled at McDonald’s Worldwide Convention in Las Vegas. It is more than a voice at the speaker box. The system greets customers, takes orders in both English and Spanish, processes modifications, displays order totals, and directs drivers to pickup windows. A video circulating on X showed the system managing a full drive-thru interaction in real time, handling language switches and order changes without missing a beat.
There is also a recognition feature. The system can identify repeat customers and suggest their usual orders, although McDonald’s has not explained how that function works technically. Still, the feature points to a broader push toward restaurant AI operations that go beyond simple order taking.
What ArchIQ does beyond the drive-thru speaker
What separates ArchIQ from a straightforward AI ordering tool is its dual role inside the restaurant. Beyond customer-facing interactions, the system monitors operations in the background, detects problems, and sends real-time alerts to managers before issues escalate.
Those alerts can cover a range of problems, from a freezer malfunction to a kitchen bottleneck slowing down order fulfillment. A franchisee account described ArchIQ as “a master brain to help managers run a better restaurant,” not just a replacement for the person holding a headset.
That operational layer matters. In a high-volume McDonald’s location, equipment failures and kitchen slowdowns can cascade quickly. As a result, a system that surfaces those problems in real time, rather than waiting for a manager to notice, adds value beyond order automation alone.
How McDonald’s is testing ArchIQ with Google infrastructure
Scale, hardware, and the Google partnership
The five-location pilot is backed by serious infrastructure. McDonald’s built ArchIQ in partnership with Google, and restaurants across the U.S. are reportedly being equipped with Google Edge Cloud hardware ahead of a broader deployment. Google’s infrastructure brings compute power and edge processing capabilities to what is ultimately a fast-moving, real-world operational environment.
The choice of Google, which in turn works with NVIDIA, suggests that McDonald’s is approaching this rollout differently than its previous AI experiment. Rather than adapting existing vendor technology, the company appears to have built something more closely aligned with its specific operational needs.
The 90% autonomous order completion rate is the headline metric from McDonald’s AI drive-thru ArchIQ testing so far. However, the underlying architecture may matter just as much for long-term scalability. If Google Edge Cloud hardware is being pre-installed across U.S. locations ahead of a wider rollout, McDonald’s is already laying the groundwork for something much larger than a five-store pilot.
Why the new AI rollout matters for McDonald’s Next
McDonald’s has been down this road before. The company spent years testing AI ordering technology with IBM across more than 100 restaurants, and that program ended in 2024 after a string of well-documented errors. One widely circulated incident involved the system adding over $250 worth of chicken nuggets to a single customer’s order, the kind of mistake that spreads quickly on social media.
That experience was not just an embarrassment. It was a data point. McDonald’s walked away from the IBM pilot with a clearer picture of what AI ordering actually requires to work at scale inside a real restaurant environment: consistency, reliability, and the ability to handle edge cases without producing absurd outputs.
The Google-backed ArchIQ approach reflects those lessons. CEO Chris Kempczinski framed it directly in terms of the company’s McDonald’s Next growth strategy. “We can’t ask our customers to choose,” he said. “Hospitality or speed.” The implication is straightforward: the right technology should deliver both rather than forcing a trade-off.
Public reaction, at least online, has been split. Some users welcomed faster, frictionless ordering, while others said they still prefer a human voice at the drive-thru window. That divide is not unique to McDonald’s; it runs through nearly every consumer-facing AI deployment right now.
What makes the ArchIQ rollout worth watching is less about whether customers like talking to an AI and more about whether the system holds up operationally at scale. The IBM pilot failed not because the concept was wrong, but because execution fell short. If ArchIQ’s 90% completion rate and real-time monitoring prove durable across more locations, McDonald’s may have found the version of drive-thru AI that actually works — and the rest of the fast-food industry will be paying close attention.
English (US) ·