Determinants and consequences of trust in AI-based customer service chatbots

ABSTRACT

According to industry reports, artificial intelligence-based chatbots could transform online customer service. Though businesses are increasingly implementing chatbots to automate customer service, the lack of consumer trust and acceptance continues to cause worry. Although trust is critical to acceptance, research on the drivers and consequences of trust in AI-based chatbots is limited. Hence a study was conducted to identify the antecedents of consumers’ trust in text-based customer service chatbots and examine the influence of trust on behavioral intentions. The data collected from 221 users was analyzed using the structural equations modeling method. Results reveal that conversational cues influence the perceived functional and social attributes of the chatbot, and these, along with personal disposition to trust technology, further influence trust formation. Finally, trust determines behavioral intentions. Incidentally, privacy risk turned out to be a non-significant predictor of trust. The study provides measures to improve trust and suggests directions for future research.

摘要

根据行业报告，基于人工智能的聊天机器人可以改变在线客户服务。尽管企业越来越多地实施聊天机器人来自动化客户服务，但消费者信任和接受度的缺乏继续引起担忧。尽管信任对于接受至关重要，但对基于人工智能的聊天机器人信任的驱动因素和后果的研究是有限的。因此，进行了一项研究，以确定消费者信任基于文本的客户服务聊天机器人的前因，并检查信任对行为意图的影响。使用结构方程建模方法分析了从 221 个用户收集的数据。结果表明，对话线索会影响聊天机器人的感知功能和社交属性，而这些，连同个人对信任技术的倾向，会进一步影响信任的形成。最后，信任决定行为意图。顺便说一下，隐私风险被证明是一个不重要的信任预测指标。该研究提供了提高信任的措施，并为未来的研究提出了方向。

KEYWORDS:

• Artificial intelligence
• chatbots
• trust in technology
• privacy

關鍵詞:

• 人工智能;聊天机器人；相信技术；隐私

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was partially supported by the University Grants Commission, India, under the Junior Research Fellowship (JRF) schemes [Grant Nos: F.15-6 (DEC. 2015)/2016 (NET) and 2917/(NET-JAN2017)].