Organizations are rapidly shifting customer interactions to artificial intelligence (AI). What is less visible is what is lost in the process of automation. At the same time, they can unwittingly weaken the customer experience and increase inequality. The more customer interaction is automated, the more likely it is that a genuine human interaction will become an additional, paid service.

Organizations automate customer interactions at an accelerating pace. This is not driven by a single technological breakthrough, but by economic logic: as digital technologies become cheaper and more capable, organizations have stronger incentives to reorganize routine work around automation (Autor 2015, 11). Rule-based AI is particularly suitable for this, as it can handle routine tasks quickly, consistently and at significantly lower costs than humans (Huang & Rust 2021, 32).

This matters as a trend because efficiency rarely remains a neutral technical objective. Over time, it shapes service models, customer expectations, and organizational ideas of what counts as “normal” interaction. The issue is not that automation fails to work, but rather what gets left out and how easily those losses remain unnoticed. From a futures perspective, the automation of customer interaction can be understood as more than a technological trend. It may also be interpreted as a weak signal: an early indication of a potentially significant societal change that is still emerging and therefore easy to overlook (Hiltunen 2012).  Seen in this light, the growing reliance on automated customer service may point toward a future in which human attention becomes increasingly scarce, segmented, and economically valued. The relevant question is therefore not only whether AI can make customer service more efficient, but also what kind of future this pursuit of efficiency gradually normalizes.

Artificial intelligence works only when problems can be clearly defined

AI works well when tasks can be clearly defined. Beyond that, its limitations become visible. The background is Polanyi’s paradox: people often know more than they can put into words. Human situational awareness, intuition and the ability to read situations are largely based on tacit knowledge that is difficult to turn into code. (Autor 2015, 11.)

This is clearly visible in customer service. Simply knowing that the service provider is AI can reduce customers’ willingness to engage with it (Luo, Tong, Fang & Qu 2019, 1). Simple requests are handled smoothly, but as the situation becomes more complicated, a human is needed. It is precisely these situations that often determine what the customer relationship will be like. At worst, the overemphasized use of AI in the customer interface can lead to customer alienation. (Huang & Rust 2021, 47.)

Human service does not disappear – it becomes scarce

When organizations optimize costs, a conflict emerges between efficiency and human interaction, alt-hough it is not always recognized. As cost optimization advances, routine interactions are increasingly handled by AI, while human interaction shifts from a standard feature to a scarce resource. It will not dis-appear – it will simply become something you pay extra for. (Huang & Rust 2021, 47.)

One possible outcome is a two-tier service future. In this future, routine cases are directed to automated channels by default, while personal service is reserved for premium customers, complex cases or those able to pay for faster access. Another, more desirable scenario is that AI is used to handle routine tasks while human support remains accessible when empathy, interpretation or trust is needed. The difference between these futures is not technological. It is strategic and political.

At the same time, AI can reinforce existing inequalities. If systems are built on biased data, they may dis-proportionately affect vulnerable groups, such as low-income individuals or heavy users of public ser-vices. These groups often accumulate more data and are therefore subject to closer scrutiny. Technical fixes alone are not enough if the underlying structures remain unchanged. Even if algorithms are improved and data is refined, the problem persists if the same biases and practices continue to shape what kind of data is collected and how it is used. (Tuominen & Snell 2025, 25–26.)

Finnish Universities of Applied Sciences involved in setting the direction

Truly sustainable development requires the ability to identify biases in AI and understand their impacts. In practice, this means strengthening AI and data literacy and being able to critically evaluate the use of technology. (Tuominen & Snell 2025, 31–34.)

The question is not just about technology, but about the choices people make when using it. This is especially relevant in higher education, where students are expected to develop the generic competences required in an increasingly digital working life. Finnish Universities of Applied Sciences have recognized this challenge and have actively sought to shape the responsible use of artificial intelligence. The Finnish Rectors’ Conference of Finnish Universities of Applied Sciences (Arene) recommendations emphasize that higher education institutions should not merely enable the use of AI, but also ensure that it is used responsibly, ethically, transparently, and in a manner that supports learning and the development of working-life competences. Special attention is given to data protection, fairness, equality, and the development of AI literacy among both students and staff. (Arene 2026.)

Laurea University of Applied Sciences has adopted a similar approach. According to its AI policy, the use of artificial intelligence is generally permitted and encouraged as support for learning, provided that its use is transparent and disclosed. At the same time, Laurea emphasizes critical AI literacy, ethical reflection, and responsible data practices, highlighting that AI should complement rather than replace human learning, guidance, and professional judgment. (Laurea 2026a; Laurea 2026b.) In this sense, Finnish Universities of Applied Sciences as active labor market partners, are not only adapting to technological change but also actively shaping the norms, competencies, and values that will guide the role of AI in future working life.

From weak signal to strategic choice

From futures perspective, the growing automation of customer interaction should be treated as a weak signal that deserves attention. It points toward a future in which efficiency, access and equality may begin to pull in different directions. If organizations only follow the logic of cost reduction, human interaction may gradually become a premium feature. If they make more conscious choices, AI can instead support better service by freeing human employees for situations where judgement, empathy and trust matter most.

The future of customer interaction is therefore not decided by AI alone. It is shaped by the values, service models and governance choices built around it. Without critical reflection, we may quickly find ourselves in a situation where empathy and understanding are no longer part of ordinary service, but a separately priced privilege.

References

• Arene. 2026. Arene’s recommendations on the use of artificial intelligence for universities of applied sciences. Updated in April 2026. Helsinki: Arene ry. Retrieved 15.7.2026.
• Autor, D. H. 2015. Why Are There Still So Many Jobs? The History and Future of Workplace Automation. Journal of Economic Perspectives 29 (3), 3–30. Retrieved 25.3.2026.
• Hiltunen, E. 2012. Matkaopas tulevaisuuteen. Helsinki: Talentum.
• Huang, M.-H. & Rust, R. T. 2021. A strategic framework for artificial intelligence in marketing. Journal of the Academy of Marketing Science 49 (1), 30–50.
• Laurea. 2026a. Laurea University of Applied Sciences. “Laurea’s AI Policy in Its Entirety.” Student’s Guide. Retrieved 15.7.2026.
• Laurea. 2026b. Laurea University of Applied Sciences. “Use of AI in Studies.” Student’s Guide. Retrieved 15.7.2026.
• Luo, X., Tong, S., Fang, Z. & Qu, Z. 2019. Frontiers: Machines vs. Humans: The Impact of Artificial Intelligence Chatbot Disclosure on Customer Purchases. Marketing Science 38 (6), 937–947.
• Tuominen, P. & Snell, K. 2025. Artificial intelligence and gender equality. Advisory Board for Gender Equality. Retrieved 25.3.2026.

Use of artificial intelligence: ChatGPT was used to support translation, language editing and structuring. The content and conclusions are the authors’ own.