Survey highlights AI use in customer service

A survey of 435 customer service professionals examined how artificial intelligence impacts their work, finding that AI is mostly used for writing and searching for information, but its concrete benefits are still limited. The first edition of the Observatory of the Impact of AI in Customer Relations, published by Smart Tribune and the blog Sens du Client, shows that while AI tools are now part of daily routines for many customer service advisors, their effects vary depending on the individuals surveyed.

Three main AI uses stand out

Three AI functions are most commonly used by customer service professionals: helping to answer questions with suggestions and reformulations, automatically summarizing conversations after calls or chats, and using an intelligent knowledge base. These tools are used by between 55.4% and 65.5% of participants, depending on how much time they spend with customers. While team productivity and work methods are viewed positively, with scores around 5.5 to 5.8 out of 10, the actual benefits are lower, with scores ranging from 2.25 to 2.56 out of 5.

Useful AI focuses on daily tasks

According to the report, customer service advisors primarily need tools that help them handle requests more efficiently. Thomas Gouritin, an AI expert at Smart Tribune and coordinator of the Observatory, explains that useful AI provides accurate answers at the right time without disrupting the conversation. Its reliability depends more on the quality of the knowledge it uses than on the technology itself. Useful AI handles routine tasks such as searching for information, categorizing and routing requests, and drafting summaries, allowing advisors to focus on customer interactions.

Frontline advisors see fewer benefits

Frontline advisors, who spend more than 75% of their time directly with customers, report fewer benefits from AI. More than half of the respondents (53.8%) fall into this category, and they give lower scores than others on all five measured benefits. On average, they score 2.26 out of 5, compared to 2.44 for other participants. In this group, the lowest score is zero, and more than one in ten advisors say AI has no impact on their work. Their use of AI is more varied, with some using it frequently and others rarely or not at all.

AI use varies by role and context

The report notes that the nature of direct customer interactions influences how AI is used. Frontline advisors are more likely to use AI for intelligent request routing (47.1%) and personalized responses (48.7%) compared to other groups. Thomas Gouritin observes that the gap in AI use between frontline advisors and others reflects a mismatch between how AI tools are designed and how tasks are performed in practice. In frontline roles, advisors handle unexpected situations in real time, and AI that is not well integrated can add complexity instead of helping.

Human skills remain vital

Despite AI's growing presence, human skills are still seen as essential in customer service. One in five respondents (20.3%) values AI for freeing up time to better serve customers. Others are concerned, with 17.8% expressing rejection or strong concerns, especially about the risk of being replaced by AI. Nearly three out of four respondents (74.5% of frontline advisors and 74% of others) cited data security and confidentiality as a major concern. In contrast, only a small percentage (8.1% and 10.6%) mentioned resistance to AI from teams.

Priorities differ by role

The priorities for AI use differ depending on the level of customer interaction. Advisors with less direct contact with customers are more concerned about the cost of AI solutions and issues like errors, biases, or hallucinations. Frontline advisors, however, are more focused on maintaining human relationships, with 45.5% worried about dehumanization and 20.9% concerned about protecting specialized knowledge. These concerns are echoed by Katia Grenu, director of customer relations at Alltricks, who emphasizes that AI cannot replace skills like negotiation, reassurance, and emotional understanding.

Training is essential for AI integration

For companies, the integration of AI into customer service requires strengthening human skills and training. The report notes that customer service professionals rate the support provided for AI tools at an average of 5.16 out of 10, with highly polarized opinions. Thierry Spencer, author of the blog Sens du Client and supporter of the Observatory, argues that AI should be treated as a human resources issue, involving both skill development and employee motivation.

AI could enhance, not replace, human roles

Thomas Gouritin notes that as AI handles routine tasks, customer service advisors will take on more central roles, focusing on sensitive and emotional issues. Advisors will also play a key role in ensuring the quality of AI, identifying errors and improving company knowledge. However, the risk is that they could be reduced to monitoring machines. The opportunity, he says, is to recognize and value their role in customer relationships and ensure they receive the recognition they deserve.