![\"Friendly](\"https:\/\/www.europesays.com\/ai\/wp-content\/uploads\/2026\/07\/1782915442_253_0x0.jpg\")
<\/p>\n<\/p>\n
Partner or Servant? Your AI Needs to Know.<\/p>\n
getty<\/p>\n
There’s an irony involved in modernizing customer communication platforms: Handing off some interactions to AI requires understanding people better than ever.<\/p>\n
The best human agents naturally adjust how they respond to each customer. Their speech, intonations and mannerisms change based on cues they pick up instinctively. They sense a customer\u2019s interest, concern, comfort level, and more. All that information helps them guide the customer to the most satisfying outcome.<\/p>\n
But AI doesn\u2019t learn the same way people learn. It needs to be taught. So now, humans need to diagnose all these elements more scientifically. Knowledge that previously seemed intuitive must be laid out in an instructive way.<\/p>\n
This is a big reason I keep an eye on the latest discoveries<\/a> about what makes AI receptionists succeed or fail. Each finding can serve as another guidepost that my team can build into the communication tools we create.<\/p>\n I\u2019ve written previously about how science is helping show what makes an agentic voice persuasive<\/a>, and how empathy is a double-edged sword — too much can put customers off. Now, it\u2019s time to look at another aspect of this technology: how friendly it should be.<\/p>\n \u2018Subordinate\u2019 vs. \u2018Social Partner\u2019<\/p>\n Customers obviously do not want an agentic tool to come across as rude. But \u201cfriendliness\u201d can also connote a sense of familiarity. That\u2019s where the challenge lies. <\/p>\n A new study published in the journal Behavioral Sciences shows that in certain contexts, customers reject AI voices that behave like a peer. They prefer a more professional style. \u201cFrom a social cognitive perspective, interaction style does more than signal friendliness; it provides information about relational norms, role expectations, and appropriate conduct, which guide how linguistic cues are interpreted,\u201d author John Yang writes<\/a>. <\/p>\n The difference involves how a customer perceives the role of AI in the interaction. \u201cFormal tone is more effective when AI is construed as subordinate, whereas casual tone is less effective under hierarchical role frames.\u201d Put simply, is the customer thinking of the AI receptionist<\/a> as a partner or a servant? <\/p>\n Cash and Conversation<\/p>\n Some people may be more apt to view AI in one way or the other in general. They may feel more comfortable treating an agentic voice in a familiar fashion, or may want to keep more emotional distance. But Yang — who got his Ph.D. at the University of Iowa and is now at the Hankuk University of Foreign Studies in Seoul — found that the context of the interaction itself is pivotal. \u201cTwo cues are especially relevant for human judgment: whether access to the AI is framed as paid or free, and whether the interaction structure is one-way or two-way.\u201d<\/p>\n Why is that? Paying for access \u201cheightens hierarchical expectations by framing the AI as an entity under the user\u2019s control or employment,\u201d he explains. So when someone accesses the AI agent for free, they\u2019re more likely to welcome a personal tone.<\/p>\n The structure of the interaction has similar effects. When a customer is looking for a sense of one-way communication, they\u2019re more likely to expect \u201casymmetry and subordination.\u201d But if they\u2019re looking for a two-way exchange, they\u2019re probably up for \u201cfostering mutuality and perceived partnership,\u201d the study explains.<\/p>\n Both of these exemplify the kinds of lessons that agentic AI tools can be taught, so they\u2019re able to recognize cues and adjust responses to deliver the best customer experience.<\/p>\n Useless Without Unification<\/p>\n Making all this work requires both depth and breadth of customer information. Think of depth as involving rich, detailed, contextual information about each customer; breadth, meanwhile, involves having robust data about as many customers as possible.<\/p>\n To walk the friendliness vs. formality tightrope and avoid leaning too far to one side or the other, a business needs to understand what a customer is looking for. That comes from pulling together every piece of data it has about the individual, across every channel, throughout their entire journey. The organization also needs to know what kind of balance works best in general for customers in similar situations. <\/p>\n