We are currently observing inconsistent formatting in chatbot responses generated from our Word-based knowledge documents. Bullet lists render differently across answers, headings are inconsistently bolded, and formatting syntax appears literally in responses. At first glance, this may appear to be a cosmetic issue. In practice, it is a customer experience, accessibility, and credibility issue.When that structure is inconsistent, the burden shifts to the customer to interpret and reorganize the information themselves. For example, when content appears as uneven hyphens or flattened text, the customer must slow down and reread. When headings appear as literal formatting symbols rather than clearly distinguished sections, the response feels unfinished or system-generated rather than authoritative. Even when the content is technically correct, the presentation can erode trust.Digital experiences are judged in seconds. Customers do not differentiate between a formatting artifact and a broader system instability. Stray asterisks or inconsistent layouts can create the perception that the chatbot is unreliable or poorly maintained. That perception increases the likelihood that customers will re-ask the question for confirmation, abandon the interaction, or escalate to a live agent. The accessibility implications are even more significant. When formatting appears as literal characters such as “asterisk asterisk heading asterisk asterisk,” a screen reader may read those symbols aloud. This adds cognitive noise and interrupts comprehension.The chatbot is a public-facing extension of our brand. If responses appear inconsistently formatted or contain visible markup artifacts, it suggests a lack of content control between knowledge management and conversational delivery. As we expand AI-enabled services, maintaining presentation quality is essential to maintaining credibility.