AIMB – AI Media Business

The TripleA research project, conducted in collaboration with Porsche Consulting, focuses on the optimization of multi-agent systems in a corporate context. Key research focuses include the orchestration of AI agents, context management, and the systematic evaluation of the developed systems. In addition, the project investigates how collaboration between AI agents and employees can be structured to increase acceptance while simultaneously improving the quality of work outcomes. The research is being conducted directly within a specific corporate environment by two research assistants under the academic supervision of Prof. Dr. Jürgen Seitz and Prof. Dr. Florian Landkammer. In the scientific analysis of the results, Hochschule Offenburg is collaborating with the research group led by Prof. Dr. Dominik Kanbach at the HHL Leipzig Graduate School of Management. At the same time, the project aims to lay the methodology and technological foundations for future innovations and product developments based on AI agents. Further areas of research include the use of so-called “model councils” to improve the quality of results, as well as the development of iterative feedback and learning loops that enable the agents to optimize their performance with increasing autonomy.

LearnSphere is an empirically oriented research project designed to investigate the acceptance and intention to use an AI-powered tutor in the form of an interactive avatar within digital learning environments for small and medium-sized enterprises (SMEs). The learning platform and the AI tutor are being developed by a project partner from the business sector, while the participating research group is responsible for the independent scientific design, implementation, and evaluation of the study. The goal of the project is to address a scientific research question regarding the perception, acceptance, and use of AI-based avatar technologies in the context of corporate continuing education. The theoretical foundation is based on established models of technology acceptance, in particular the Technology Acceptance Model (TAM) and its extensions, which are empirically tested for the application context of AI-supported learning systems. In addition, constructs such as perceived comprehensibility, motivation, and subjective learning success are integrated into the research model. The empirical investigation is conducted using standardized quantitative surveys of specialists and managers from SMEs and is evaluated using a hypothesis-driven approach. The results contribute to the academic literature on the acceptance of AI-supported learning systems and enable the derivation of generalizable insights for the deployment of such technologies in the SME context.

KISEE (AI-Based Skills Assessment and Development) is a collaborative research project led by mycelia gGmbH in Berlin and the Institute for Applied Artificial Intelligence at the Stuttgart University of Media. This research project aims to investigate how an AI-based advisor can help ensure equal opportunities in accessing future skills. People from

households with limited educational opportunities and those with an immigrant background are expected to benefit particularly from this initiative. An interactive AI avatar is at the heart of the project and is designed not only to complement traditional counseling approaches such as chatbots or employment agencies but also to make them more accessible, scalable, and effective. The avatar offers personalized guidance, assesses existing skills, opens up new opportunities in the job market, and facilitates access to continuing education programs and certifications. Its effectiveness is being evaluated in three areas: In comparison with chatbots, the study examines whether using the avatar leads to more precise skill profiles and greater information gain. The Technology Acceptance Model (TAM) is used to assess acceptance and willingness to use the avatar. In addition, independent experts are reviewing the accuracy of the skills analysis performed by the interactive AI avatar. The goal is to make access to future-relevant skills more equitable and to create sustainable development opportunities for everyone.

The ValueTales research project examined how generative AI can help companies make data-driven decisions more quickly and with greater confidence. At the heart of the project is the concept of “Smart Executive Summaries”—clear, easy-to-understand reports automatically generated by a chatbot that translate complex KPI data into concrete recommendations for action, thereby serving as a digital decision-making partner. The accompanying study clearly demonstrates how AI chatbots support specialists and executives in decision-making processes and under what conditions they are accepted and used in everyday business operations. Based on an empirical study involving over 100 participants, the findings show that perceived added value is the key driver for adoption. AI is used primarily when it enables better, more precise, and immediately actionable decisions. At the same time, trust in data sources, transparency of results, and the quality of responses play a central role in acceptance. The findings also make it clear that time savings and efficiency gains are perceived as significant competitive advantages, and that intuitive use and low cognitive load facilitate adoption. Furthermore, organizational factors such as leadership, corporate culture, and digital competencies significantly influence adoption. ValueTales thus provides concrete guidelines for companies on how to effectively design and sustainably establish AI-powered decision support.

TALAINTED is a collaborative research project between the Hacker School in Hamburg and the Institute for Applied Artificial Intelligence (IAAI) at Stuttgart University. The goal of the project was to investigate how AI-supported learning formats can sustainably inspire young people—particularly girls and socioeconomically disadvantaged groups—to pursue careers in the IT and digital industries during the career orientation phase. As part of the project, an innovative, multilingual online course was developed that relies on the use of virtual AI avatars. These avatars serve as digital learning guides and enable flexible, self-directed learning—regardless of location, time, or individual learning pace. Through step-by-step instructions, interactive elements, and unlimited opportunities for repetition, low-threshold access to programming content was created. The effectiveness of the approach was examined in a study involving 272 students aged 12 to 14. The results show that the AI-based learning format achieves significantly positive effects: Interest in programming increased markedly, as did the motivation to engage more deeply with digital topics. Participants also reported improved concentration, an easier understanding of the content, and increased enjoyment of learning. Interest in a future career in the IT field was also sustainably fostered. Furthermore, important insights were gained regarding the design of inclusive and effective digital learning offerings. In particular, it became clear that AI avatars can help reduce educational barriers and better support individual learning processes. A key outcome of the project is the development of a practical guide that helps schools and organizations independently implement AI-based learning formats. The insights and concepts gained through the project will be specifically shared to enable other educational stakeholders to develop comparable programs. The online course is available at the following link: https://hackersc.uber.space/talainted/

TALAINTED impressively demonstrates how artificial intelligence can be used in education to promote equal opportunities, establish innovative learning formats, and guide young people on their path toward a digital future.

In a world where education and technology are becoming increasingly intertwined, a fascinating question arises: Can semantic artificial intelligence revolutionize learning through educational videos? This question was the focus of the AIEDN (AI Education) project by the Institute for Applied Artificial Intelligence at HdM Stuttgart, thingsTHINKING GmbH, and math YouTuber Daniel Jung. After a year of intensive research, the project’s conclusion offers exciting insights into the future of learning. At the heart of the project was the development of a semantic AI learning assistant. This advanced tool understands learners’ questions on a semantic level and provides targeted answers from an extensive pool of educational content. It enables learners to formulate questions in everyday language and be directed straight to relevant sections in learning materials. As part of the project, researchers investigated whether the AI learning assistant not only makes learning more efficient and motivating but also whether it can facilitate long-term learning success and a deeper understanding of the topics covered. In addition, potential applications of AI learning assistants in corporate settings were explored. A particular priority for the project participants is to use the AI video assistant to provide all learners, regardless of their social background, with equitable access to education and individualized support, thereby taking an important step toward educational equity. The AI learning assistant can be used free of charge at www.aiedn.com.

Integrated Instruction on the Responsible Use of Artificial Intelligence Based on Physical-Virtual Demonstrators

Artificial intelligence (AI) has become a key technology of the 21st century, permeating many aspects of our work and daily lives. Given this reality, there is a growing need to incorporate AI more fully into higher education. However, course offerings have often been limited to a single perspective—most commonly that of Informatik. The IKID project addressed this challenge with an integrated and innovative curriculum.

IKID: An Integrated Approach to AI Education
IKID pursued the goal of providing a holistic education for the AI experts of tomorrow. AI was examined not only from the perspective of Informatik, but also from the perspectives of

Betriebswirtschaft, ethics, and law. This multi-perspective and practice-oriented approach aimed to equip students with all the necessary skills to implement AI projects successfully and responsibly. This qualification was certified with the “Interdisciplinary AI Competencies” certificate.

Development and Evaluation: The Interdisciplinary Didactic Concept
The IKID project encompassed the development, as well as the scientific research and evaluation, of an interdisciplinary didactic concept. This concept defined the methodology and content-related framework for a holistic AI education. Through continuous evaluation, the IKID team ensured that the curriculum met current requirements and was continuously optimized. The result is a field-tested and scientifically sound model for an interdisciplinary AI teaching concept.

AI Education: Multi-Perspective and Practice-Oriented
The courses began with fundamentals in Informatik, economics, ethics, and law to equip students with the tools for AI projects from their respective disciplines. The focus was on understanding the interests of the participating disciplines, leveraging synergies, and resolving conflicting goals. HdM students were able to regularly enroll in the interdisciplinary AI program starting in the 2023–24 winter semester.

Physical-Virtual Demonstrators: Hands-On
AI In addition to the introductory courses, the project team relied on the use of physical-virtual demonstrators. These brought AI to life for the students and provided opportunities for interaction and discussion. Thematically, the demonstrators represented various application areas of AI and gave students the opportunity to directly apply what they had learned. As part of integrated AI projects, for example, students took on the role of senior staff members—as in a case study—represented the interests of the participating disciplines throughout the process, and ultimately brought their project to a holistic success. In an AI sandbox developed specifically for the project, students were provided with a secure environment in which they could train and test AI algorithms themselves and develop their own projects.

Outlook: An Open Offering for All Universities
Upon completion of the project at the end of 2025, the IKID team made the developed materials and methods available as an “open-source solution.” Other universities can benefit from this and integrate the interdisciplinary AI exploratorium into their own teaching concepts. With the IKID project, the HdM took a decisive step toward forward-looking, interdisciplinary higher education that optimally prepares students for the challenges and opportunities of AI.

The research project “What can AI do for me?” was launched in 2021 by the Institute for Applied Artificial Intelligence at the Stuttgart Media University (HdM), thingsTHINKING GmbH, and KENBUN IT AG. The goal of the project was to promote specialization in the field of digital value creation and to build bridges between businesses and the ever-evolving AI landscape. Following intensive research, the project launched the AI-based matching platform “WhatCanAIDoForMe.com” in October 2021. Here, companies can describe their projects or challenges in text form using everyday language. The semantic matching system then identifies suitable AI applications and potential AI solution providers. At the same time, users receive an expert assessment of the value-creation potential of comparable

projects. This is intended to help companies discover the diversity of AI applications and find the ideal project partner. The findings from the AI Value Creation study form the core of the platform. This study examines the multifaceted aspects of AI’s value creation potential in companies. In over 40 qualitative interviews, more than 90 AI use cases were identified and analyzed. The research findings shed light on both the challenges companies face when implementing AI and the broad spectrum of value creation potential that AI initiatives enable across all areas of the business.