U.S. consumers are ready to dine out again. With rising vaccination rates, nationwide fast-food chains are flush with customers wanting to eat out after a year at home. But with many employees still unable or unwilling to return to work, many companies have turned to AI to increase productivity, enhance customer service, and stay competitive.
Consider McDonald’s, which started using speech recognition software to take more orders at drive-throughs in Chicago, a service that saw huge spikes in demand during the pandemic. Or Burger King, which uses AI to suggest items on its food board, such as iced drinks on a hot summer’s day instead of hot ones. Or White Castle, the American hamburger chain, which is using an autonomous frying chef in 10 locations to optimize the cooking of each component in a customer’s order so everything is ready simultaneously (thereby doing away with cold, limp fries).
But despite the many benefits AI brings to consumers and the economy, some groups are trying to use the court of public opinion to demonize the use of the technology by fast-food companies. News coverage has focused on the purported ill-effects of AI, sounding the alarm about big fast-food companies manipulating consumers, having data breaches, and frustrating customers. But policymakers should resist giving in to calls that would put the brakes on technological innovation.
The primary concern of some opponents is that fast-food chains will use AI to manipulate unsuspecting consumers into purchases that undermine their interests. For instance, McDonald’s acquired an AI tool called Dynamic Yield in 2019 to identify repeat customers and personalize its drive-through menu for them, which one article argues is heralding in the “McData-fueled future of capitalism.” Corporations, these commentators claim, are analyzing data to turn a profit by unfairly keeping consumers “coming back for more no matter what.” This claim might make sense if McDonald’s was engaging in nefarious or anti-consumer behavior, like quietly tacking on additional charges to a customer’s order. But that does not seem to be the case. On the contrary, personalizing their service in a consumer-friendly way to keep repeat customers coming back is a central part of the business model for all fast-food chains, and frankly, virtually every business that sells directly to consumers. In fact, according to a recent McKinsey report, 80 percent of consumers expect and desire personalized experiences throughout their customer experience.
Still, the misguided view that consumers are passive sheep being manipulated into turning over their valuable data and hard-earned cash can lead policymakers to enact unnecessarily strict laws on data collection and use, such as affirmative consent laws. The problem with these laws is that they saddle companies with unnecessary red tape and limit beneficial uses of data. For example, Illinois’ Biometric Information Privacy Act allows consumers to sue organizations for processing their biometric data without their express written consent. Unfortunately, this law has resulted in a significant, frivolous lawsuit against McDonald’s for its speech recognition technology at drive-throughs, which only serves to disincentivize franchisees in the state from adopting AI that benefits consumers.
Another major concern is that as large fast-food chains collect large amounts of consumer data for AI, they become bigger targets for hackers looking to exfiltrate sensitive consumer data. Indeed, there have been a slew of high-profile data breaches at companies like Wendy’s, Arby’s, McDonald’s, and Chipotle, and brands have had to determine how to improve their cybersecurity practices, including among individual franchises. But hackers are not interested in what kind of sauce consumers want for their nuggets or what size slurpee they bought. The primary target of hackers is financial data, such as credit card numbers, and attacks usually target the stores’ point-of-sale systems, which are often provided by third parties. Fast-food companies will need to address cybersecurity threats regardless of whether they adopt AI.
The final concern is that inaccurate AI systems may frustrate consumers, making the overall customer experience worse. For example, some commentators have implied that McDonald’s speech recognition software will annoy customers because it is only 85 percent accurate and 20 percent of orders need human intervention. But they overlook that humans aren’t much better. A 2019 study of drive-through performance with humans found that the average order was 84 percent accurate because menus are getting more complicated, lanes are getting busier, and there is more pressure on employees to work quickly.
If anything, AI offers the opportunity to make the fast-food customer experience less frustrating by becoming more inclusive for people with disabilities and those who do not speak the local language. For instance, franchisees that use AI-enabled self-order kiosks or mobile apps can offer locally adapted menus that make it easier for the nearly 8 million Americans who have problems with speech. And some kiosks, like those Juke Slots offer, can help people who are deaf or have hearing loss by displaying a virtual avatar that communicates in sign language. Encouraging restaurants to adopt such solutions can reduce the reams of incidents where fast-food companies refuse to serve people who are deaf. Besides, in a free-market economy regulators don’t need to tell companies how to improve customer service; they are already plenty motivated to do so on their own.
Overall, succumbing to AI opposition is near-sighted folly. Only by de-stigmatizing the technology can policymakers begin to have constructive conversations about what they should do to adequately capture the technology’s benefits while protecting consumers.
Image credits: Flickr user.