Netizens gang up on AI Drive Through and order 18,000 cups of water

The age of artificial intelligence has arrived at the drive-thru window, but it’s not quite the smooth ride that restaurants expected. Taco Bell’s ambitious AI experiment has hit some major speed bumps, thanks to creative customers who’ve discovered just how easy it is to outsmart the technology.

Since launching AI-powered drive-thrus at over 500 locations in 2023, the fast-food giant has processed more than two million orders through their voice recognition system. The goal was simple: faster service and fewer mistakes. What they didn’t anticipate was the internet’s endless creativity when it comes to testing technological limits.

The breaking point came when social media exploded with videos of customers exploiting the AI’s vulnerabilities. In one viral clip, a user managed to order an absurd 18,000 water cups, effectively crashing the entire system.

This wasn’t an isolated incident. Other viral videos showed the AI stumbling through basic interactions, like asking customers about drink preferences immediately after they had already placed their beverage order. These embarrassing moments quickly spread across social media, turning what was meant to be a technological showcase into a source of entertainment for pranksters.

The backlash has forced Taco Bell to pump the brakes on its AI rollout. Dane Mathews, the company’s Chief Digital and Technology Officer, acknowledged the challenges in a candid assessment of the program’s performance.

“I think like everybody, sometimes it lets me down, but sometimes it really surprises me,” Mathews told The Wall Street Journal. “We’re learning a lot.”

Rather than abandoning the project entirely, Taco Bell is pivoting toward a more nuanced approach. The company has recognized that AI isn’t a universal solution that works equally well in all situations. During peak hours at busy locations, human employees often prove more effective at managing the chaos of long drive-thru lines.

The new strategy involves a hybrid model where human staff members can monitor AI interactions and intervene when the technology falters. “For our teams, we’ll help coach them: at your restaurant, at these times, we recommend you use voice AI or recommend that you actually really monitor voice AI and jump in as necessary,” Mathews added.

While the company maintains that voice AI remains part of their long-term strategy, the exact role of this technology is still being determined.