{"slug": "starbucks-quietly-retired-its-ai-agent-just-months-after-deployment-after-it-and", "title": "Starbucks quietly retired its AI agent just months after deployment after it hallucinated coffee shop inventories and slowed down baristas", "summary": "Starbucks has quietly scrapped its AI-powered inventory management system after only nine months, following reports that the tool frequently miscounted or mislabeled items and slowed down baristas. The coffee giant confirmed to Fortune it has moved back to a single model of counting inventory, as the NomadGo-provided app often \"hallucinated\" inventories by failing to identify bottles on shelves. The decision comes as part of CEO Brian Niccol's \"back to Starbucks\" turnaround plan, which has shown early success with a 7.1% increase in quarterly comparable U.S. sales.", "body_md": "[Starbucks](https://fortune.com/company/starbucks/) has quietly scrapped its AI-powered inventory management system after only nine months since its deployment.\n\nThe coffee giant confirmed to *Fortune* it has made an operational decision to move to a single model of counting inventory following an announcement in September to deploy its automated counting tool.\n\nThe app, provided by NomadGo, took inventory of beverage components like milk and syrups in order to keep track of item shortages. In February, *Reuters*, which [first reported](https://www.reuters.com/business/starbucks-scraps-ai-inventory-tool-across-north-america-2026-05-21/) the discontinuation of the tool this week, cited Starbucks sources who said the app often miscounted or mislabeled items, “hallucinating” inventories by failing to identify the presence of bottles on shelves.\n\n“We test ideas in our coffeehouses, listen closely to partner feedback, and make changes to deliver a better, more consistent experience,” a spokesperson told *Fortune* in a statement.\n\nNomadGo didn’t immediately respond to *Fortune*’s request for comment.\n\nCarl Addison, a Starbucks shift supervisor of nine years based in Shoreline, Wash., told *Fortune *the automated counting app required stores to rearrange back-of-house storage, which was a time-intensive process. The app’s inaccuracies made employees’ workflow more challenging, he said. If the system counted too much of the product, it wouldn’t send enough of a product a store was running low on. If the system counted too little, it wouldn’t ship enough of a needed product.\n\n“It started off not particularly accurate and got less accurate over time,” Addison said.\n\nStarbucks sent *Fortune* a handful of barista responses to the automated counting tool expressing that it improved inventory processes and the interface to view inventories.\n\n“Thanks for discontinuing Automatic Counting! The thought behind it was great, but the execution was proving difficult,” one comment read.\n\n## Brian Niccol’s ‘back to Starbucks’ plan\n\nStarbucks has implemented a host of AI tools as part of its “[back to Starbucks](https://fortune.com/2026/01/30/starbucks-ceo-brian-niccol-turnaround-plan-measuring-stores-five-metrics/)” plan under CEO Brian Niccol to improve slumping sales and streamline operations. The coffeehouse’s recent AI deployments include [Green Dot Assist](https://fortune.com/article/how-is-starbucks-using-ai-in-coffee-shop-locations-ceo-brian-niccol-green-dot-assist/), an app on the store’s iPads able to provide recipe cards and appropriate ingredient substitutions, as well as troubleshoot issues with machinery. It has a Smart Queue tool that sequences orders to improve order speed and efficiency. Former CEO Laxman Narasimhan said in early 2024 customers were [abandoning mobile orders](https://fortune.com/2024/05/01/starbucks-earnings-morning-commute-app-orders/?utm_source=search&utm_medium=suggested_search&utm_campaign=search_link_clicks) because of long wait times and product availability.\n\nSo far, Starbuck’s turnaround strategy, which also includes adding cozier seating and paring back menu items, [appears to be working](https://fortune.com/2026/04/29/starbucks-coo-turnaround-workers-baristas-stores-pay-benefits/). Last month, the company [reported](https://fortune.com/company/starbucks/earnings/q2-2026/) a 7.1% increase in quarterly comparable U.S. sales last quarter, beating analysts’ expectations of a 4.5% increase. Quarterly revenue increased 9% to $9.5 billion.\n\n## Retail’s automation challenges\n\nStarbucks’ decision to revert back to its previous inventory system also reflects broader growing pains in how the retail industry has deployed AI. Earlier this month, a major Pizza Hut franchisee [sued the chain](https://fortune.com/2026/05/19/pizza-hut-franchisee-lawsuit-ai-adoption-doordash-delivery-drivers/) over its AI program. The franchisee claimed Pizza Hut’s Dragontail Artificial Intelligence system gave gig workers increased visibility to internal systems that enabled them to leverage AI systems for their own benefit, like selecting orders with larger tips and bunching orders, resulting in delayed deliveries and “cascading operational breakdowns.”\n\nAs global restaurant automation is expected to balloon into a [$28 billion market](https://restaurant.org/education-and-resources/resource-library/make-room-for-restaurant-automation/) this year, the pressure is on for these technologies to deliver.\n\nAt this point of AI’s development, the challenges retail spaces are facing in scaling the technology has led Santiago Gallino, a Wharton professor of operations, information, and decisions, to this conclusion: “Right now, there is more hype than actual benefit.”\n\n“Many retailers feel the pressure to say they are doing AI-related things and AI-related innovations and running these things before they’re ready to give concrete and real returns,” he told *Fortune*.\n\nGallino applauded Starbucks’ decision to walk back its use of the automated counting tool. Inventory management is an ongoing issue in retail, he said, and while technology has advanced to allow companies to improve non-trivial inventory challenges, optimization tools are not a panacea to these issues.\n\nOther companies like Zara, have spent years trying to refine its use of algorithmic technologies. The fast-fashion retailer [implemented a microprocessor-based tagging system](https://www.reuters.com/article/business/zaras-tagging-system-means-even-faster-fashion-idUSKBN0FK1NI/) more than a decade ago, tagging inventory with Radio-Frequency Identification (RFID) that ultimately improved the accuracy of inventory and made it easier to track items across its system.\n\nAccording to Gallino, the Zara case study is less an argument about technology generating universal benefits to retailers, but rather an example of a company doing the research and iterating a technology’s use to fit its specific needs. While the onus is on retailers to leverage budding technology, AI as a whole will only become a sustainable technology for these companies if it offers a return on investment.\n\n“One general theme that to me is still a little bit perplexing, is how, on many levels, [return on investment] seems to be not a main consideration—the promise that down the road all this is going to make sense,” Gallino said. “That is something that can be out of focus in the middle of the hype.”\n\nAddison, the shift supervisor, said at this juncture, barista’s workflows are not best served by the technology.\n\n“I would love AI if I felt like it worked, but have to say…I just don’t feel like it’s a solid fit for a retail environment, where accuracy and speed are both really important,” he said. “And it just doesn’t feel like it can really deliver on those fronts for us.”\n\n**The Fortune 500 Innovation Forum** will convene Fortune 500 executives, U.S. policy officials, top founders, and thought leaders to help define what’s next for the American economy,\n\n**Nov. 16-17 in Detroit.**", "url": "https://wpnews.pro/news/starbucks-quietly-retired-its-ai-agent-just-months-after-deployment-after-it-and", "canonical_source": "https://fortune.com/2026/05/28/starbucks-quietly-retired-ai-inventory-agent-barista-complaints-hallucinations/", "published_at": "2026-05-28 19:22:59+00:00", "updated_at": "2026-05-28 19:58:50.006698+00:00", "lang": "en", "topics": ["artificial-intelligence", "ai-products", "ai-tools", "ai-agents"], "entities": ["Starbucks", "Fortune", "NomadGo", "Reuters", "Carl Addison"], "alternates": {"html": "https://wpnews.pro/news/starbucks-quietly-retired-its-ai-agent-just-months-after-deployment-after-it-and", "markdown": "https://wpnews.pro/news/starbucks-quietly-retired-its-ai-agent-just-months-after-deployment-after-it-and.md", "text": "https://wpnews.pro/news/starbucks-quietly-retired-its-ai-agent-just-months-after-deployment-after-it-and.txt", "jsonld": "https://wpnews.pro/news/starbucks-quietly-retired-its-ai-agent-just-months-after-deployment-after-it-and.jsonld"}}