GLOSSARY
Anthropomorphic AI
Anthropomorphic AI appears human through language, voice or design; this impression establishes neither consciousness nor a technical architecture modelled on the brain.
A familiar tone changes the interaction
A friendly AI response can make a tool easier to approach. In sales, this creates an important question: is a negotiation suggestion accepted because of its substance, or because it feels like advice from an experienced colleague? A persuasive tone can conceal uncertainty when sources and assumptions remain invisible. This matters particularly when the answer confirms the user’s own view. Agreement alone provides no evidence that the opposing position has received serious consideration.
Connect trust to verifiable performance
An assistant’s design should make its capabilities, limitations and responsibilities understandable. Statements such as “I know your customer” may imply more knowledge than the available context supports. References to specific documents, unanswered questions and alternative interpretations are easier to assess. A user study can present the same recommendation in different tones: does acceptance change although the evidence remains identical? Such observations help distinguish useful accessibility from excessive trust. Explainability contributes when it reveals the actual basis for a decision rather than merely making the system sound reassuring.