TLDR: Taco Bell is re-evaluating the widespread deployment of its voice AI ordering system in drive-thrus after experiencing mixed results, including glitches, order inaccuracies, and customer discomfort. Despite processing over 2 million orders, the company’s Chief Digital and Technology Officer, Dane Mathews, acknowledges the technology’s inconsistent performance and suggests that human intervention might be more effective during peak hours, indicating a more strategic and selective approach to AI integration moving forward.
Taco Bell is undertaking a significant reassessment of its artificial intelligence (AI) powered drive-thru ordering system, acknowledging that the technology’s performance has been inconsistent across its more than 500 installed locations. Dane Mathews, Taco Bell’s Chief Digital and Technology Officer, openly admitted to The Wall Street Journal, ‘We’re learning a lot, I’m going to be honest with you. I think like everybody, sometimes it lets me down, but sometimes it really surprises me.’
The fast-food giant’s venture into AI for order-taking has been met with a spectrum of reactions, ranging from surprising successes to widespread customer frustration and even viral online ‘trolling’ incidents. Since its integration last year, the AI system has processed approximately two million orders, yet it has been plagued by glitches, delays, and a notable inability to accurately process complex or accented orders.
Customers have reported comical and costly errors, including instances of bacon being added to ice cream and charges for thousands of dollars due to the AI adding dozens of extra items. One particularly notable incident involved a customer successfully ordering 18,000 cups of water, which required human intervention to cancel. Another customer even managed to order McDonald’s items through Taco Bell’s AI, with the system surprisingly accepting the order and suggesting competitor menu items.
These challenges highlight a growing gap between the promise of AI in controlled environments and its messy reality in real-world customer service. Mathews indicated that the company is now considering a more strategic deployment, suggesting that human staff might be more effective during peak hours at busy restaurants where long lines are common.
This reconsideration by Taco Bell is not an isolated incident within the fast-food industry. McDonald’s, for example, concluded a two-year AI drive-thru experiment with IBM in June 2024, though it has since partnered with Google Cloud to continue exploring the technology. Yum! Brands, the parent company of Taco Bell, KFC, and Pizza Hut, also announced a strategic partnership with Nvidia in March, aiming to deploy various AI solutions, including drive-thru systems.
Also Read:
- The Evolving Role of Call Center Agents: AI as an Enhancer or a Replacement?
- Generative AI’s Creative Edge Over Humans Remains Unproven, Meta-Analysis Suggests
Despite the setbacks and the ‘AI shame’ phenomenon reported in workplaces where executives and Gen Z employees hide their AI usage due to lack of training , Taco Bell still plans to advance with AI voice technology. The current experience is shaping the company’s understanding that deploying AI exclusively at high-volume locations may not always be the optimal approach, emphasizing the need for a balanced integration of AI and human capabilities.
