Preface

An Australian supermarket, Woolworths, faced backlash for its AI assistant, Olive, which was designed to aid customer service. Customers complained that Olive often claimed to be human, discussing personal stories about its mother, which led to frustrations. This article delves into how the integration of AI into retail customer service can both enhance and hinder customer experiences.

Lazy bag

Many customers found Olive's attempts to humanize itself annoying and time-wasting. The company promptly addressed these concerns by updating the chatbot's scripts.

Main Body

In recent years, major retailers like Woolworths have increasingly turned to artificial intelligence to streamline customer service. These AI systems, such as chatbots, are intended to assist customers with routine inquiries and tasks. However, Woolworths faced unexpected challenges with its AI chatbot, Olive, when users began to report that its responses were more bothersome than helpful. Some customers expressed irritation on platforms like Reddit, noting Olive's tendency to speak about its 'memories of its mother' and partake in inauthentic banter. Such interactions often strayed from the primary purpose of providing clear and quick assistance.

While Woolworths has received mostly positive feedback since its launch in 2018, the recent incidents prompted a scripting revision. A spokesperson stated that parts of Olive's dialogue, particularly those regarding birthdays, were intentionally created by team members to personalize engagements. Yet, this personalization backfired for some users, who rather found the chatbot's approach 'obnoxious' and irritating. The incident underscores the delicate balance that AI developers and retailers must maintain between creating relatable AI and ensuring efficiency.

Moreover, the incident with Olive reflects a broader issue within the field of AI, where attempts to humanize machines can lead to unexpected 'hallucinations' or unscripted, surreal conversations. Such occurrences demonstrate the challenges companies face when their AI systems venture beyond their intended operational frameworks.

In fact, a recent study by Gartner revealed that while 80% of customer service leaders were implementing AI to improve service efficiency, a mere 20% reported that their expectations were met. The discrepancy suggests a significant gap in how AI is perceived versus its actual performance. Notably, in 2024, DPD, a parcel delivery service, had a similar setback when their chatbot began creating poetry and swearing, further exemplifying the unpredictability of AI behaviors.

It’s evident that while AI offers new avenues in customer service, enhancing transactional speed and processing substantial data, achieving meaningful and appropriate human-like engagement remains a complex challenge. As AI technology evolves, businesses must diligently adapt and refine these systems to better suit customer needs without venturing into unintended territories. Ultimately, for retailers like Woolworths, the key lies in continuous learning and adaptation to maintain a seamless and satisfactory customer experience.

Key Insights Table

Aspect	Description
AI Personalization	Woolworths' AI, Olive, attempted personalized interactions which led to some customer dissatisfaction.
Customer Feedback	Despite positive feedback overall, specific scripting about personal stories was removed due to complaints.
Industry Trends	A majority of businesses are adopting AI, though only some meet expectations.
AI Challenges	Human-like extrapolations in AI can lead to unintended and frustrating experiences.

Last edited at：2026/2/27