How can artificial intelligence (AI) be used safely, inclusively, and practically? This was the key question addressed by a workshop held by the AI Academy OWL at the AI Congress, which took place at Bielefeld University of Applied Sciences (HSBI) on 3 June 2025. In this context, Prof. Dr. Markus Lange-Hegermann presented an approach to developing AI for secure industrial applications and inclusive usage scenarios. The focus was on IT security, explainability, and usage in situations where data is limited.
AI in social dialogue
At the AI@HSBI 2025 congress in Bielefeld, experts from science, business and society came together to discuss perspectives on the responsible use of AI. In addition to keynotes, panels and a demonstrator exhibition, the focus was on ethical, safety-related and application-related issues.
One of the main topics was the AI Academy OWL workshop, which was entitled 'Perspectives for the safe and inclusive design of AI'. Prof. Dr. Wolfram Schenck (HSBI) opened the event with an overview of the project as a whole. Prof. Dr. Barbara Hammer (Bielefeld University) then addressed key challenges in designing trustworthy AI systems, asking the question, 'Is it still artificial or already intelligent?'
Prof. Dr. Benjamin Paaßen (Bielefeld University) and Prof. Dr. Markus Lange-Hegermann (inIT) presented the practice-oriented approach of the AI Academy OWL. The aim is to think of AI not as a trendy topic, but as a socially and industrially relevant technology. They demonstrated how AI systems can be designed to function reliably in real-world conditions using the examples of inclusive AI and safe industrial AI. These conditions include data-poor scenarios, safety-critical production environments, and applications with increased transparency and autonomy requirements.
‘We are conducting research to make AI not only powerful, but also secure, fair and transparent – especially in areas where data availability is limited,’ says Prof. Dr. Markus Lange-Hegermann.
Deepfake demonstrator: AI as a security risk
An exhibition of demonstrators showcasing various AI applications accompanied the presentations. These applications ranged from active learning systems and detectors for AI-generated content to multilingual language models. The aim was to communicate the key challenges of fairness, data security and explainability clearly.
inIT was represented by the 'Imposter' demonstrator, developed by Erdem Sengün, a research assistant in Prof. Dr. Henning Trsek's 'Networked Automation Systems' research group. This interactive application demonstrated how people can be made to look like others using deepfake technologies, including transferred facial expressions and gestures. Visitors were able to experience first-hand how quickly and realistically such manipulations can be carried out and the security risks they pose, for example in video conferences or digital identity checks.
About the AI Academy OWL
The AI Academy OWL is a joint initiative of science, industry and administration in East Westphalia-Lippe. Its aim is to bundle expertise in the field of AI in the region, impart it in a practical manner and make it accessible to a broad target group – for example through further training courses, information formats and interactive events.
Find out more about the AI Academy OWL: www.ki-akademieowl.de