McDonald’s targets AI ordering at most U.S. restaurants by the end of 2027

McDonald’s expects AI ordering to reach most of its U.S. restaurants by the end of 2027, with the company projecting substantial labor savings. But some franchisees question whether the gains will cover the costs—and whether customers want to give their orders to software. The chain is now widening trials of its Archy ordering system.

CEO Chris Kempczinski outlined the expansion during Wednesday’s earnings call, Restaurant Dive reported. Additional drive-thrus in the United States and Australia will test Archy, and McDonald’s plans further international pilots throughout 2027. The timetable sets out two different stages: expected deployment across most U.S. restaurants by year-end, alongside continued testing in other markets during the year.

Taking the order, freeing up staff

Archy includes technology that takes spoken drive-thru orders. It sits within ArchIQ, the artificial intelligence platform McDonald’s announced in September under its Next initiative. That platform also includes automated inventory tracking and connected equipment, putting customer-facing order-taking alongside systems intended to help run the restaurant behind the scenes.

At the chain’s September investor day, executives said Archy could take orders in English and Spanish and had accuracy above 90%. They also described customer satisfaction as high during early tests. Those company-reported results provide the performance backdrop for the wider trials; the planned 2027 deployment remains an expectation, not a completed rollout.

McDonald’s expects the technology to free up at least 50 labor hours each week. Kempczinski’s stated aim is to let employees concentrate on customers and food quality. The company is presenting automated order-taking as a way to redirect staff attention toward other restaurant work, with the labor-hour figure describing an expected benefit.

While Archy appears to be making progress on accuracy in tests and franchisees are eager for the labor savings (indeed it's needed to justify the investment…), some questioned if the consumer actually wants this technology, and if voice AI is inconsistent with the renewed focus on service and hospitality

Peter Saleh, restaurant analyst, in an October report quoted by Restaurant Dive

The savings have to survive the bill

The financial ambition extends beyond the ordering speaker. Technology is part of McDonald’s plan to improve restaurant margins by 250 basis points, or 2.5 percentage points. Saving labor hours could contribute to that goal, but the margin target covers a broader technology effort. It is not a projected return from Archy alone.

Franchisee concerns predate the latest trial expansion. In an October report, Peter Saleh, a U.S. Bancorp | BTIG managing director and restaurant analyst, wrote that operators had received little detail about specific productivity improvements. According to Restaurant Dive’s account of those concerns, the margin target seemed “highly theoretical” to operators. Their objection was to the case being made for the gains, not simply the use of AI.

Operators also wanted a fuller accounting of expenses across the systems. Saleh cited technology fees and radio-frequency identification tags among their cost concerns. Some questioned how much of the effort was genuinely new. Accounting for both the benefits and the bills, they argued, would leave a smaller margin improvement than the headline target suggested.

Their strongest enthusiasm was for inventory management, which Saleh described as a smaller part of the margin target. Franchisees saw opportunities to save managers’ time, reduce food waste and keep less inventory on hand. That preference distinguishes their assessment of the back-office tools from their more cautious reception of an AI system that speaks directly with customers.

Kempczinski said McDonald’s would regularly update investors on progress and results from key elements of Next throughout 2027. Those promised updates will accompany the wider trials and expected U.S. deployment.