Shoppers will soon have the right to request human assistance over an AI chatbot, under new EU rules being implemented by the Department of Finance.
The rise of AI chatbots has undeniably streamlined operations for online retailers, offering cost-saving efficiencies and quicker responses. Yet, many consumers have found themselves frustrated with these automated systems. Repetitive answers and a lack of nuance have led to a growing chorus of shoppers wishing for the human touch. Soon, they’ll have the right to bypass chatbots altogether and speak to a real person when they need assistance.
On Lunchtime Live, listener Garrett expressed his clear preference for human interaction, stating that resolving issues is often more straightforward when speaking directly to a person. “I always feel that you can work through the details and get to the nub of the issue faster than working with an AI system,” he explained. “They may have their place, but once you start getting complicated or outside of the norm, they have a tendency to just break down and fail.”
In the same programme, TU Dublin lecturer Damian O’Reilly discussed the evolution of chatbots. “Traditional chatbots were rule-based—they didn’t adapt to changing questions and gave the same scripted answers,” he noted. While AI chatbots are now smarter, he cautioned that they still lag behind well-trained humans. O’Reilly referenced a study from Oxford that highlighted the limitations of AI; when working with precise information provided by experts, a chatbot achieved a 95% accuracy rate. However, when patients self-reported their symptoms, that accuracy plummeted to just 35%. “So, it depends on what you’re putting into it is what you’re getting out of it,” he added, reminding us all that human intuition still plays a critical role in customer service.
In our view, one might argue that while AI can be a helpful tool, it should never fully replace the human element in customer care. Yet we’ve all been there—staring at an endless chatbot loop, feeling more exasperated with each “Let me redirect you!” than the last. Maybe it’s time we introduce a scale for chatbot frustrations, ranging from mild annoyance (a brief chuckle) to “I’ve just spent 20 minutes discussing my issue with a computer.” What’s next? Chatbots that draft poetry about our grievances? Here’s hoping those remain just fanciful dreams.
Did You Know? The first chatbot, ELIZA, was developed in the 1960s and was designed to simulate a conversation with a psychotherapist. It might have been basic, but it paved the way for the sophisticated AI we see today—reminding us of just how far technology has come (or how far it still has to go) in understanding human needs.
