Figure 4. Screenshot showing ChatGPT in a further attempt to provide real references, once again unsuccessfully.

Our experience with ChatGPT was very positive and its support was helpful in writing this case, especially for the general microbiological information of Pasteurella multocida and the aspects related to the debridement, antibiotics, and implant retention surgical procedure. However, we found a limitation of ChatGPT in providing reliable and actually existing literature references. This phenomenon, called “hallucination,” has already been described. Deep learning-based generation is prone to “hallucinate” unintended text, which degrades the system performance and fails to meet user expectations in many real-world scenarios.