A 311 program is evaluating a vendor's claim that its new resident chatbot is an "AI agent." Which characteristic would actually support that claim, rather than describe a simple scripted decision tree?
Select an answer to reveal the explanation.
Short Explanation
A scripted decision tree is like a choose-your-own-adventure book: every path was already written down before the resident ever called. An agent is different because it decides on its own, in the moment, which tool to reach for, which nobody scripted step by step in advance.
Full Explanation
The defining trait of an AI agent is autonomy paired with tool use: the system decides for itself which action or system to invoke to accomplish a goal, rather than following a path a person pre-wrote for every possible situation. When a 311 chatbot autonomously chooses to query a case-status system because that's what the resident's question requires, it's exercising the kind of independent judgment that separates an agent from a script. A fixed script with a predetermined menu path is the textbook definition of scripted logic, not agentic behavior, since every branch was decided in advance by a developer. Requiring the resident to select from a numbered list before the system does anything describes menu-driven interaction, where the system has no real decision-making role at all. Pre-written replies that never vary regardless of input describe a static FAQ bot, which has neither autonomy nor tool use. Before accepting the vendor's "AI agent" label, the 311 program should ask for a concrete example of the chatbot choosing between two different backend systems on its own for two similarly worded resident questions, since a vendor that can't produce that example is likely describing a decision tree with agent branding.