“You often want to talk to me during customer service calls.” Thanks to my CX analyst, this clue from the recent New York Times mini-crossword puzzle gave me a pause. I knew that “human” had to be the answer… despite the industry's best efforts to insert AI agents into customer interactions.
When studying business and consumer perceptions of AI's use in CX, Metrigy discovered several areas where the two do not match. Here is one example: Most companies using AI (75.9%) feel that customer service has improved last year, according to a 1,104 global AI for Business Success 2025-26 survey. However, consumers are divided into opinions on the quality of customer service interactions. A survey of 503 North American consumers in April 2025 showed only 35.6% reported better customer service. Others were pretty evenly divided, not seeing change and seeing deterioration in customer service.
Furthermore, a majority of consumers (almost 70%) expect that AI will have at least a massive (25.6%) impact on future interactions with companies (44.1%), but not everyone has a good impact. We expect the impact of AI to be negative, a significant percentage of 41.1%.
One potential Din may come when consumers encounter AI text and voice agents deployed by many companies to promote customer self-service. Consumers using AI agents nod to the improvements they see, but the overall customer experience is inconsistent. 26.4% of consumers experienced “significant variability” using AI agents, and 55.3% noted that they encountered “some degree of variability” in AI-based interactions.
Also, AI agents are not yet a reliable service option for many consumers. 38% of consumers said they would avoid using AI text agents, compared to just 19.5% of consumers who prefer AI text agents. 26.5% of customers avoid voice AI agents, and 30.5% of customers prefer voice agents. The main complaint that leads to customers avoiding both text and voice AI agents is that agents can't understand their needs, and customers stay in the AI-Agent interaction loop without a clear path to humanity. Conversely, time savings and convenience are the biggest reasons why AI agents are used. The key to a company is that if they don't fully understand the needs of consumers or provide a simple escape route to live agents, AI implemented to improve efficiency will generate more frustration than convenience.
Generic AI is widely used, but our research shows that trust is limited among consumers. Compared to 60.8% who only trust AI in selected situations, or 18.2% who trust AI at all, or 21.0% who don't trust it at all, there are only 21.0% who are completely reliable. And as much as companies are talking about making AI more and more autonomous, 28.4% of consumers show that AI never trusts them to complete tasks on their behalf. The main concerns of consumers are concerns about data privacy (57.6%), ethical use of data (47.3%), malicious use of technology (46.4%), and production, including information accuracy (40.9%).
Building Trusts deprive you of transparency and provide practical value. According to the consumers surveyed, if AI is being used (46.1%), they should ensure human monitoring of AI content creation and decisions (41.2%) and provide assurances that personal information is not being used in ways that compromise privacy (39.5%).
Beyond trust, businesses can increase the value of CX's AI to suit consumer preferences. For example, 39.2% of consumers appreciate at least some personalizations during the interaction. This includes using preferred channels and conversation tones.
CX leaders are totally enthusiastic about AI possibilities, but many consumers are more cautious. Building trust and demonstrating concrete benefits will help CX leaders in winning more consumers into AI courts.
