McDonald's Drive-Thru AI Upgrade: What You Need to Know About the Rollout
McDonald's Drive-Thru AI Upgrade: What You Need to Know About the Rollout
From IBM to Google: A Shift in Strategy
McDonald’s first experimented with AI drive-thru ordering in 2021, launching a pilot program with IBM that expanded to more than 100 restaurants. The technology, known as Automated Order Taker (AOT), had its share of growing pains. Videos of the system going viral on TikTok showed AI making odd mistakes — adding items customers didn’t request, suggesting bacon with ice cream, and even taking orders for 18,000 cups of water.
At its peak, the voice-recognition technology was accurate about 85% of the time, though human staff had to intervene in roughly one in five orders. McDonald’s ended its partnership with IBM in 2024, but made it clear that voice ordering remained part of its future plans. “As we move forward, our work with IBM has given us the confidence that a voice ordering solution for drive-thru will be part of our restaurants’ future,” the company said at the time.
Enter Google Cloud. McDonald’s entered into a partnership with Google in late 2023 to integrate edge computing technology across its global locations. The shift allows each restaurant to process data locally rather than relying on distant cloud servers — a critical advantage for locations with unreliable internet connections.
Introducing ArchIQ: The New AI Operating System
The centerpiece of McDonald’s latest AI push is ArchIQ, a voice-activated system currently being tested at five U.S. locations. Early demonstrations show Archy — as the system has been nicknamed — accurately taking orders in both English and Spanish. According to a franchisee account on X, more than one million transactions have been processed through the system, with roughly 90% completed without requiring human escalation.
But ArchIQ is more than just an order-taker. The system is designed to function as what one franchisee described as a “master brain” for restaurant operations — alerting managers to potential bottlenecks or equipment issues before they disrupt service. Google’s edge computing technology enables smart devices throughout the kitchen, from fryers to ice cream machines, to transmit real-time data for predictive maintenance.
The company plans to install Google Edge Cloud hardware at every McDonald’s in the U.S. in anticipation of a broader rollout. McDonald’s CEO Chris Kempczinski has framed the initiative as part of the “McDonald’s Next” growth strategy, aimed at “improving unit economics” while bringing in more customers.
Beyond Voice: Accuracy Scales and Ready on Arrival
Voice AI isn’t the only technology McDonald’s is deploying. The chain has introduced AI-powered “Accuracy Scales” that weigh each bag of food before it’s handed to customers. If the weight doesn’t match the expected total for the items ordered, staff are alerted to double-check the contents before the customer drives away.
The company is also expanding its “Ready on Arrival” program, which uses geofencing technology in the McDonald’s mobile app to detect when customers are approaching. The system alerts kitchen staff so they can begin preparing the order in advance, reducing wait times. During an August earnings call, Kempczinski said restaurants with the system can cut pickup wait times by more than 50%, and in many cases eliminate them entirely.
McDonald’s is also adding multiple drive-thru lanes at many locations, creating additional capacity to handle increased volume. The company expects these physical and technological upgrades to be largely completed by 2027.
Mixed Consumer Response and Ongoing Challenges
Despite the efficiency gains, customer reception to AI drive-thru ordering remains decidedly mixed. Social media reactions following the ArchIQ announcement included comments like “No one wants this — we like dealing with smiling faces,” with others expressing concern that automation is making human interaction increasingly rare.
The challenges aren’t just perceptual. A 2025 Drive-Thru Study by Intouch Insight found that AI-enabled drive-thrus were the fastest, with a service time of 3 minutes and 53 seconds, beating the overall average of 4 minutes and 15 seconds. However, AI locations scored lower on friendliness — 72% compared to 78% at traditional locations — and order accuracy actually lagged slightly, at 83% versus 87% at non-AI locations. Customizations were the biggest problem, accounting for 62% of incorrect AI orders.
Yet when employees stepped in to support AI systems, accuracy jumped to 95% — suggesting that a hybrid human-AI approach may be the most effective path forward.
McDonald’s technology head, Durga Prakash, said the company plans to “double down” on AI investments by 2027, with ambitions to expand AI order verification to 40,000 locations globally. The company is already using AI to verify orders at 400 restaurants, catching errors before food is handed to customers.
Conclusion
McDonald’s ArchIQ represents one of the most ambitious AI deployments in the fast-food industry, combining voice ordering, edge computing, accuracy verification, and predictive maintenance into a single integrated system. While early testing suggests real improvements in speed and operational efficiency, the company still faces significant hurdles — not least of which is convincing customers that AI can deliver the friendly, reliable service they expect.
For now, the Golden Arches are moving forward with AI, learning from past missteps while betting that technology, thoughtfully deployed, can elevate both customer satisfaction and the bottom line. Whether American consumers will embrace Archy at the speaker box — or continue to long for a human voice on the other end — is a question that only time, and millions of drive-thru transactions, will answer.
