Should You Label AI as AI?
Yes. You should always label an AI support agent as AI. Disclosing that a customer is talking to a bot is a core trust practice: it sets expectations, reduces backlash when something goes wrong, and is increasingly expected by customers and regulators. Hiding the AI to make it seem human is a short-term trick that erodes trust the moment the seam shows, and it can create liability when the bot makes a mistake the customer took as an official human answer.
Why disclosure is a trust multiplier
Customers are already skeptical of support bots, with Gartner reporting that many are more likely to reach for a third-party AI tool than the company bot. Pretending a bot is human does not fix that skepticism, it weaponizes it. Clear labeling, paired with a fast and accurate answer, is what actually rebuilds trust, because it lets the quality of the resolution speak for itself.
Labeling done well
Good disclosure is honest without being apologetic: name the AI, make its status obvious, and make the path to a human clear. The Cursor incident, where an undisclosed-feeling bot answer drove cancellations, shows the cost of ambiguity. The fix the company itself landed on was to clearly label AI responses. Transparency plus a clean escalation path is the pattern.
How Resolve handles it
Resolve treats "always label the AI as AI" as a standing design rule, and Reva reads sensitive context such as account, tier, history, and sentiment before answering. The AI is presented as an AI operator, not a human in disguise, and any ticket it should not handle escalates to a person through the decision card. Trust comes from honesty about what the customer is talking to, plus the quality of the answer.