For many readers of popular media, the coming AI revolution might seem inevitable. Nearly constant reports on large corporations’ plans to develop their own AI products suggest that in the near future, consumers will enjoy unparalleled efficiency and service provision by flawless technology. But as the punchline to an old joke goes, “Not so fast, Johnson!”

The problems associated with AI are starting to dominate news stories, replacing the previously laudatory claims, even for the tech giants that have functioned as the primary champions and earliest adopters of AI. For example, Google quickly introduced AI-generated answers in its search engine results with fanfare, but the answers offered prompted a lot more questions. One user reported that an AI-generated search result advised them to eat glue. Another received inaccurate information about Barack Obama, including a misidentification of his religion.

Search results are not the only outcome being potentially undermined by AI technology. McDonald’s has been partnering with IBM to add technology to its drive-thru lanes that can process customers’ orders and potentially even provide personalized recommendations for regular customers. Instead of enhancing efficiency though, the AI-enabled drive-thrus appear prone to a wide range of errors, which consumers also have publicized widely. In videos that appear equally frustrating and hysterical, social media users reveal their confusion when they receive ketchup packets instead of ice cream or try to juggle nine beverages being handed to them, when they only ordered two.

Likely in response to the rapid and expansive sharing and circulation of these errors, public backlash has been swift. Some consumers have sworn off companies that use AI-supported technology (or at least those that make their usage evident). Other market actors call for regulatory changes and updated laws to monitor AI-generated content and impose meaningful sanctions for not just incorrect fast food orders but also fake news that threatens more detrimental societal implications.

In turn, many of the companies that publicly adopted AI applications also have publicly announced how they would be adjusting their uses. For example, Google immediately rolled back some AI features, while promising to undertake much more testing and stricter quality control measures before reintroducing it. McDonald’s paused its uses of AI systems in drive-thrus and indicated that its partnership with IBM would focus instead on how to use technology to achieve more internal operational efficiency. But the damage to public trust and their brand reputations already has been done. The question for the future is whether and how AI applications can be improved, to ensure and exploit the substantial potential benefits. Artificial intelligence is not going away; for companies and brands, the goal must be to find ways to use it strategically, to accrue all its benefits while avoiding its potential detriments.

Discussion Questions

1. What are the risks of implementing advances but still inaccurate technology for (a) companies and their reputations, (b) consumers and users, and (c) society as a whole?
2. Considering your answers to the preceding question, should AI implementation be limited by regulations, or should it continue without hindrance, so that the applications can learn and improve?

Sources: Gerrit De Vynck, “Google Scales Back AI Search Answers After It Told Users to Eat Glue,” The Washington Post, May 30, 2024; Chris Morris, “McDonald’s Just Fired its Drive-Thru AI and Is Turning to Humans Instead,” Fortune, June 17, 2024; Jessica Apotheker, Sylvain Duranton, Vladimir Lukic, Nicolas de Bellefonds, Sesh Iyer, Olivier Bouffault, and Romain de Laubier, “From Potential to Profit with GenAI,” BCG, January 12, 2024; Phillip Shinall, Elodie Teboul, Francesco Bellino, Matt Gamber, Shannon McDonald, and Clement Richet, “How Consumer Goods Companies Win in Turbulent Times,” BCG, February 21, 2024; OpenAI ChatGPT, “Assistance with Research on AI Implementation Challenges in Tech and Consumer Goods Companies,” ChatGPT, July 7, 2024