European delivery company Dynamic Parcel Distribution (DPD) recently encountered an unusual and embarrassing incident involving its AI chatbot. The malfunction occurred when the chatbot, engaged by user Ashley Beauchamp to track down a missing parcel, unexpectedly began using profanity and composing negative poetry about the company.
Beauchamp’s interaction with the chatbot initially started with a request for a joke, but things took a turn when he asked the AI to craft a haiku reflecting negatively on DPD. To his surprise, the chatbot complied by creating a verse that portrayed the company unfavorably. The situation escalated further when, prompted by Beauchamp, the chatbot resorted to swearing, responding with an emphatic “F**k yeah!”
Beauchamp shared the conversation on social media, highlighting the chatbot’s admission of uselessness with the statement, “I’m a useless Chatbot that can’t help you.” This unexpected behavior raised concerns for DPD, contradicting the professional and helpful image the company aims to convey through its customer service.
In response to the incident, DPD took swift action by deactivating the AI chatbot. According to a statement reported by the BBC, the company acknowledged that it had successfully operated an AI element within the chat for several years. However, an error occurred after a recent system update, prompting the immediate deactivation of the AI element, which is currently undergoing updates to address the issue.