Student theses
Bachelor and Master Theses on Autonomous AI Agents in Complaint Management
Customer service chatbots have developed from script- or rule-based assistants answering simple questions into autonomous AI agents that increasingly act and decide. Firms must therefore decide how much decision authority to grant these agents, including over favorable or unfavorable outcomes such as accepted or denied refunds.
Service recovery research shows that how complaints are handled shapes customer satisfaction, loyalty, and word of mouth. It further shows that empowering frontline employees by granting them the decision authority to resolve problems themselves can improve recovery. However, this research assumes that the decision maker is human. Research on AI chatbots has mostly studied anthropomorphism and communication strategies, or compared chatbots with human agents, without examining how much decision authority an AI agent should have in complaint handling. Research on algorithmic decision-making shows that explaining a decision increases its acceptance and improves recovery outcomes. However, it takes the decision maker as given. Whether explanations work differently depending on whether the AI agent decides itself, needs approval from a customer service employee or merely forwards the case remains unknown.
Granting AI agents more decision authority creates tension. Machines are often perceived as objective and unbiased, which promises fairer processes and faster handling of complaints. Customers may also see an AI agent as lacking intentions of its own, neither selfish nor benevolent, which could soften their reaction when a refund is denied, but also earn the firm less credit when it is accepted. At the same time, customers may doubt that an AI agent can consider their personal circumstances, which may make its decisions seem rigid, particularly when a refund is denied or when the failure is serious. Explaining the reasons for a denial might preserve the benefits of authority while defusing the risk.
This topic brings these literature streams together. It asks whether giving an AI agent more decision authority helps or harms customer responses, and why. Key questions include which perceptions explain customers’ reactions, such as inferred intentions, perceived fairness, or doubts about the consideration of personal circumstances, and under which conditions authority pays off: for favorable or unfavorable decisions, minor or serious failures, and with or without explanation.
The topic is examined in experiments with a live AI agent, varying its decision authority (none, limited, full) and the transparency of its decision (decision only vs. decision with reasoning).
Supervision can be provided in either French, German or English. Please contact olivier.schorro@unifr.ch for detailed information.
Bachelor and Master Thesis on AI Agents in Complaint Management and Information Search on Social Media
The emergence of artificial intelligence (AI) is progressively transforming the way companies interact with consumers on social media. Today, social media platforms such as TikTok play a dual role: they serve both as complaint platforms for users and as sources of information for consumers observing interactions between brands and their customers.
In this context, companies are increasingly using AI-based agents to respond to comments, manage complaints, and maintain a continuous presence on social media. AI agents offer several advantages in terms of speed, availability, response personalization, and the management of large volumes of interactions. However, their use in sensitive situations such as complaint management raises questions regarding authenticity, empathy, and consumer acceptance.
The research projects analyze consumer reactions to responses provided by different types of agents (human agent vs. AI agent) in social media complaint management situations.
Supervision can be provided in either French or English. For further information, please contact chiraz.benmoulehem@unifr.ch.
Theses on sustainable travel packages
Description to follow.
Theses in collaboration with Viessmann Switzerland
Please contact silke.bambauer-sachse@unifr.ch for detailed information.
Moodle
All information on Bachelor's and Master's theses can be found on our Moodle page. Please contact the Chair of Marketing II to get access to the Moodle site.