Skip to content

Survey: How do chatbots influence attitudes and knowledge around sustainability? Take the survey (10 min)

Project ReCAIcle
Back to news

How AI Chatbots Drive Sustainable Choices

One year of ReCAIcle: 1,181 documents screened, three working hypotheses formulated, preliminary studies conducted and first results published.

Project ReCAIcle

Mockup of a sustainable AI chatbotAI generated

The ReCAIcle research project officially started on 1 October 2024. Its goal is to investigate how AI-supported chatbots can help people make sustainable and resource-conserving decisions. The aim is not to force users into anything, but to find out how digital assistants must be designed so that they provide orientation, build trust and make sustainable behaviour easier. In particular, this is intended to help companies and their employees use physical resources – such as company laptops, tools or machines – more consciously and for longer. This is to be achieved through digital impulses, so-called nudges.

Foundation work in the first half of the year

The first half of the project was all about laying the groundwork. As part of the successfully completed work package 1, a total of 1,181 relevant documents on the use of nudges in conversational agents were reviewed, evaluated and transferred into a database. This comprehensive analysis not only provided practical implementation methods, but also identified new, open research questions. The results of this literature review and analysis were summarised in a scientific paper1, which provides an overview of the current state of research, including identified key technologies, methods and application areas of AI-supported conversational agents, and is currently in the review process for ECIS 2026. The project website was also launched during this first period.

Empirical studies in the second half of the year

In the second half of the project, the focus was on empirical studies and hypothesis development. For this purpose, three central working hypotheses were formulated as part of the second work package, dealing with the effectiveness of different nudge types in promoting sustainable purchasing decisions, their efficiency, and the design of chatbots to increase user trust and encourage behaviour.

To test feasibility and validate the hypotheses, preliminary studies were finally conducted in cooperation with seventhings GmbH and TU Dresden. In an online pre-study2, for example, the design of a simulated chatbot interaction was examined to see how it can influence the choice of shipping method. The focus was on the question of whether different psychological approaches, such as social norms compared to preset default options, can more effectively encourage the decision for sustainable shipping. This study was presented in detail in January at the International Marketing Trends Conference 2026 in Berlin and subsequently published. Furthermore, a quantitative study specifically examined chatbot design, in particular the humanisation of chatbots and brands. A key result is that users’ willingness to donate depends largely on trust in the brand, while trust in the chatbot plays a comparatively smaller role3. Finally, another qualitative study showed that customer interaction with AI chatbots can be divided into different phases. From these customer experiences, concrete best and worst practices can be derived that help companies optimise the implementation of chatbots4.

In addition, the InnoTeam ReCAIcle took part in the FutureSAX Innovation Conference 2025 and published contributions on TU Dresden’s social media channels56.

In order to put the theoretical insights gained into practice, close cooperation with a practice partner is planned for the future, in order to transfer the research results into real application scenarios.

Sources

Footnotes

  1. Rüger, Theres; Janssen, Ulrike; Eulitz, Laura; Prägert, Johanna; Bellger, Mariam, “How conversational agents guide users: a systematic literature review on nudging strategies and their implementation in conversational agents” (2026), ECIS 2026. ↩

  2. Rüger, Theres; Janssen, Ulrike; Ghaheri, Negin; Merres, Sandrine; Schlegel, Vanessa; Schubert, Lukas; Werner, Friederike, “Designing Conversational Agents to Promote Sustainable Behavior” (2026), IMTC 2026. ↩

  3. Rüger, Theres; Brendel, Alfred Benedikt; Bellger, Mariam; and Mücksch, Janek, “More Than a Mascot? The Effect of Conversational Mascots on Donations for Nonprofit Organizations” (2025), ICIS 2025 Proceedings, 11. https://aisel.aisnet.org/icis2025/is_good/is_good/11 ↩

  4. Kober, Evelina; Rüger, Theres; Siems, Florian Ulrich, “Herausforderungen von KI-Chatbots in Kundeninteraktionen” (2026), Marketing Review St. Gallen, 2026(1), 42–49. ↩

  5. https://www.youtube.com/watch?v=VU_OwmOuLJE ↩

  6. https://tu-dresden.de/ing/maschinenwesen/imm/td/forschung/forschungsprojekte/recaicle ↩

All posts