Ilustration by AI
Why AI Chatbots Still Fail to Feel Human, According to New Indonesian Research
SURABAYA, Indonesia — Artificial intelligence chatbots have become the front line of customer service across major e-commerce platforms, but new research from Universitas 17 Agustus 1945 Surabaya finds that users still view them as machines rather than human communicators. The study, published in 2026, reveals a growing gap between technological sophistication and the human expectations that shape customer experiences on platforms such as Shopee and Tokopedia.
The research, conducted by Muhammad Hanif Salafuddin, Teguh Priyo Sadono, and Merry Fridha Tri Palupi of Universitas 17 Agustus 1945 Surabaya, examines how Indonesian consumers interpret interactions with AI-powered chatbots. The findings suggest that while chatbots have become faster, smarter, and more conversational, they continue to fall short of the human qualities users value most: empathy, authenticity, warmth, and meaningful interpersonal communication.
AI Chatbots Are Everywhere, but User Frustration Remains
The rapid adoption of AI chatbots has transformed customer service worldwide. E-commerce companies increasingly rely on automated systems to provide 24-hour assistance, answer questions, and manage customer inquiries at scale.
In Indonesia, chatbot technology is widely used by major e-commerce platforms. Previous studies cited in the research indicate that more than 70 percent of Shopee’s initial customer interactions are handled by chatbots, while Tokopedia’s chatbot systems manage more than half of customer inquiries.
Despite these advances, dissatisfaction remains common. International surveys have shown that many consumers still prefer human assistance over AI-driven customer service. The Indonesian study offers a deeper explanation for this disconnect by focusing on users’ lived experiences rather than technical performance metrics.
According to the researchers, the issue is not simply whether a chatbot can answer questions correctly. The more important question is whether users perceive the interaction as genuinely human.
Exploring Human Expectations Behind Digital Conversations
To understand how people experience chatbot communication, the researchers used a qualitative approach involving in-depth interviews, observation, and documentation analysis.
The study focused on three active e-commerce users who regularly interacted with customer-service chatbots on Shopee and Tokopedia. Participants included a homemaker, an entrepreneur, and a private-sector employee, all of whom had extensive experience using chatbot services.
Rather than measuring response speed or accuracy, the researchers examined how participants interpreted and evaluated their interactions with AI systems.
This approach revealed a consistent pattern: users recognized chatbot limitations even when the technology attempted to imitate human communication.
Eight Key Findings From the Research
The study identified several recurring themes that appeared across all participants.
Users Always Recognize Chatbots as Machines
Despite friendly greetings and conversational language, participants consistently reported knowing they were communicating with a machine. Structured responses, repetitive wording, and predictable interaction patterns reinforced that perception.
Chatbot Communication Feels Different From Human Communication
Participants described chatbot conversations as rigid and formulaic. Even when responses appeared polite, they lacked the spontaneity and flexibility associated with human interaction.
Expectations Often Go Unmet
Users expected chatbots to solve problems, not merely provide scripted answers. When systems failed to offer practical solutions, frustration increased.
Chatbots Work Best for Simple Issues
Participants acknowledged that chatbots were useful for routine tasks such as tracking orders, checking account information, or answering basic questions. However, performance declined when issues became more complex.
Cultural Values Shape User Judgments
The study found that communication expectations are strongly influenced by cultural norms. Values such as politeness, friendliness, empathy, and interpersonal warmth serve as benchmarks when users evaluate chatbot performance.
Simulated Empathy Is Not Viewed as Genuine
Many chatbot systems use phrases designed to sound empathetic, such as apologizing for inconvenience or expressing understanding. However, participants frequently interpreted these statements as automated scripts rather than authentic concern.
Chatbots Are Viewed as Tools
All participants regarded chatbots as practical tools rather than communication partners. They valued the convenience but did not perceive the systems as substitutes for human interaction.
Human Communication Remains Irreplaceable
The strongest finding was that participants believed chatbots would never fully replace human customer-service representatives. They argued that genuine understanding, emotional awareness, and contextual judgment remain uniquely human capabilities.
The Paradox of Humanizing AI
One of the most significant contributions of the study is its explanation of what the authors call the “humanization paradox.”
As chatbot developers make AI systems more human-like, users raise their expectations accordingly. Ironically, when chatbots fail to meet those higher expectations, disappointment becomes even greater.
Features designed to make chatbots appear human can sometimes have the opposite effect. For example, instant responses may improve efficiency, but users often interpret rapid replies as evidence that they are interacting with a machine rather than a person.
Similarly, attempts to simulate empathy can highlight the absence of genuine emotional understanding.
The researchers argue that these contradictions reveal a fundamental challenge in human-machine communication: technology can imitate human behavior, but imitation alone does not create authentic human connection.
Implications for E-Commerce and AI Development
The findings carry important implications for businesses investing in AI-powered customer service.
According to the researchers, chatbot success should not be measured solely by technical performance indicators such as speed, accuracy, or availability. Companies must also consider how users interpret communication through cultural and social expectations.
As Muhammad Hanif Salafuddin and colleagues from Universitas 17 Agustus 1945 Surabaya explain, chatbot interactions are evaluated through humanistic communication values that include empathy, authenticity, equality, and interpersonal warmth. When these expectations are not met, users perceive a communication gap regardless of technological sophistication.
The study suggests that future chatbot development should prioritize human-centered design, cultural sensitivity, and contextual understanding rather than focusing exclusively on automation efficiency.
For e-commerce companies, this could improve customer satisfaction and strengthen trust in AI-assisted services.
Author Profiles
Muhammad Hanif Salafuddin, M.I.Kom.
Communication researcher at Universitas 17 Agustus 1945 Surabaya. His research focuses on digital communication, artificial intelligence, and human-machine interaction.
Teguh Priyo Sadono, M.Si.
Lecturer and researcher at Universitas 17 Agustus 1945 Surabaya specializing in communication studies, media, and digital society.
Merry Fridha Tri Palupi, M.Si.
Academic and researcher at Universitas 17 Agustus 1945 Surabaya with expertise in communication, culture, and new media studies.
Source
Article Title: The Paradox of Humanizing AI-Based Chatbots with Humanistic Communication among E-Commerce Users
Authors: Muhammad Hanif Salafuddin, Teguh Priyo Sadono, Merry Fridha Tri Palupi
Journal: Formosa Journal of Social Sciences (FJSS)
Year: 2026
Volume/Issue: Vol. 5, No. 2, pp. 69–80
DOI: https://doi.org/10.55927/fjss.v5i2.7
Official Journal URL: https://journalfjss.my.id/index.php/fjss/index
0 Komentar