Dairy Queen just joined the AI drive-thru wave, but the real story is who's actually taking your order.
The Summary
- Dairy Queen is deploying AI chatbots from Presto to dozens of drive-thrus across the US and Canada, following a test last year
- Presto already powers voice ordering for Carl's Jr., Hardee's, Taco John's, and Fazoli's, but a 2023 Bloomberg report revealed the "AI" often includes human workers in the Philippines listening in
- The stated goal: speed up service and upsell customers on adding more items to their orders
The Signal
Dairy Queen isn't breaking new ground here. They're following a well-worn path that chains like McDonald's, Wendy's, and Taco Bell have been testing for years. What's worth noting is the vendor: Presto, a company that went public via SPAC in 2022 and has been quietly building the infrastructure layer for fast food voice automation.
The technology story is straightforward. Speech recognition captures your order, natural language processing decodes "yeah gimme that Blizzard thing with the Oreos," and the system routes it to the kitchen display. The pitch to franchisees is compelling: fewer labor hours, faster throughput, higher average ticket sizes through algorithmic upselling.
"The goal is to speed up drive-thru service and encourage customers to add more food to their orders."
But here's where it gets interesting. Bloomberg reported in 2023 that Presto's "AI" wasn't always fully autonomous. The system routes difficult orders to human workers, often based in the Philippines, who listen in and course-correct when the bot gets confused. This isn't a bug. It's the actual architecture of Web4 agents in production.
- The AI handles simple, high-volume transactions (80-85% of orders)
- Human operators catch edge cases: thick accents, custom modifications, system failures
- Customers have no idea whether they're talking to silicon or a person halfway around the world
This is what real AI deployment looks like in 2026. Not full autonomy. Not "the robots took all the jobs." It's hybrid intelligence at scale, with humans in the loop handling exceptions while machines do the repetitive work. The economics work because offshore labor costs less than US minimum wage, and the AI layer reduces the total human hours needed.
The Implication
Watch the vendor consolidation play. Presto is building the rails for an entire industry's voice interface layer. If they can prove unit economics across Dairy Queen's footprint, they become infrastructure, not a feature. That's the real bet.
For workers, this is the shape of things: not replacement, but relegation to exception handling. The question isn't whether AI takes the job. It's whether the job that remains pays enough to matter, and whether it's happening in your zip code or someone else's.