Research program: Design of trustworthy artificial intelligence
As part of the research program, principles of trustworthy artificial intelligence are developed, technically implemented and tested together with users in practical fields of application. A particular focus is on social media and the question of how they can be improved.
Trust in AI through understanding
Trust in AI through co-design
The research program follows in the tradition of participatory and user-centered software development and extends this with a focus on artificial intelligence and machine learning. Participatory software development offers numerous methods for actively involving users in software development. Our goal is to further develop these methods with regard to artificial intelligence and machine learning as a new type of co-creation.
One area in which the research program will explore methods of human-AI interaction is social media. This application context is particularly suitable as AI systems play a central role on platforms such as Facebook, YouTube, TikTok, Instagram and Twitter. Up to 70% of videos viewed on YouTube were selected by an AI system. Research into human-AI interaction is socially relevant, as users generally have no control over AI systems and are often unaware of the systems’ existence.
Trust in AI through control
Prof. Dr. Hendrik Heuer
Straight to …
Research areas
The research focus aims to measure and analyze the understanding of users of AI systems (user beliefs, folk theories). We extend preliminary work on the understanding of users of AI systems, especially video recommendation systems based on machine learning. A particular focus is on the transparency and explainability of AI systems.
This research focus develops methods for controlling the output of ML-based systems through systematic audits. An important motivation is the prevention of discrimination. The focus is motivated by preliminary work on online radicalization on YouTube.
We are also engaged in the development of methods for the comprehensive inclusion of users in the design of AI systems. Our focus on deep user involvement is based on preliminary work such as an interdisciplinary collaboration with social scientists and an extensive literature review on the different approaches to user involvement. The focus is in the context of the democratization of machine learning.
News
Why Disinformation Remains a Problem – and What We Can Do About It
On Friday, May 16, 2025, CAIS researcher Prof. Hendrik Heuer will be a guest at the Bergische Expo. In a short talk, he will address...
CAIS at re:publica 2025: Understanding Disinformation, Rethinking Content Moderation
Researchers from CAIS will participate in Europe’s leading conference on the digital society – re:publica 2025. Their focus: digital trust, platform responsibility, and algorithmic fairness.
Balancing Realism and Trust: AI Avatars in Science Communication
AI avatars are increasingly used in science communication — but what makes them trustworthy? A recent study by Jasmin Baake, Dr. Josephine B. Schmitt and...
CAIS Researcher at CHI 2025 in Japan
Prof. Dr. Hendrik Heuer and his team from the CAIS research program “Design of Trustworthy Artificial Intelligence” will be represented at the ACM CHI Conference...
Pre-CHI 2025 in Siegen
Researchers from the CAIS research program “Design of Trustworthy Artificial Intelligence” participate in Pre-CHI 2025 in Siegen from April 2 to 4, 2025.
Recommendations for Independent and Effective Disinformation Monitoring in the Digital Space
CAIS supports the open letter from the Bertelsmann Foundation, CeMAS, Democracy Reporting International and ISD Germany, which provides recommendations to the new federal government for...
The arc from technology to people
Hendrik Heuer will speak at the panel discussion “Der Bogen von der Technik zum Menschen” (The arc from technology to people) at the University of...
Hendrik Heuer Appointed as Expert Member of the NRW Inquiry Commission on Artificial Intelligence
Hendrik Heuer has been appointed to the NRW state parliament’s Enquete Commission IV, “Artificial Intelligence – For a Smart State in a Digitalized Society.”
Reenvisioning Interaction for the Age of Human-Centered Artificial Intelligence
On February 10, 2025, Prof. Hendrik Heuer will give a lecture on Human-Centered Artificial Intelligence at the University of Graz at the invitation of Prof....
Measures Against Disinformation Online
For the ARD Mittagsmagazin, we prepared a statement on online disinformation, which was included in a segment on “Fake News in Election Campaigns – A...
AI Ethics Unwrapped: The role of ethical principles in AI development
In 2024, the article “AI Ethics Unwrapped: An Empirical Investigation of Ethical Principles in Collaborative Ideation Processes” was published, examining ethical principles in the AI...
From Augustus to Trump – Why Disinformation Remains a Problem and What We Can Do About It
Josephine B. Schmitt and Hendrik Heuer will give a talk on December 29, 2024 at the 38th Chaos Communication Congress hosted by the Chaos Computer...
Workshop discussion: “A ZDF for all”
On 27 November 2024, Hendrik Heuer took part in the workshop discussion “One ZDF for all” at the ZDF broadcasting centre in Mainz at the...
6 criteria for assessing websites: Study by Hendrik Heuer and Elena L. Glassman
In a study, Prof Dr Hendrik Heuer from the Center for Advanced Internet Studies (CAIS) and Elena L. Glassman from Harvard University present six criteria...
Hendrik Heuer on the relevance of Joseph Weizenbaum’s thinking in the digital exhibition
The Weizenbaum Institute’s digital exhibition “Pioneer, Sceptic, Activist: The Turning Points in the Life of Joseph Weizenbaum” is now live. Hendrik Heuer has contributed an...
Workshop on user-centred artificial intelligence in Cagliari
Hendrik Heuer is co-organiser of “UCAI ’24: Workshop on User-Centered Artificial Intelligence” in Cagliari, Sardinia, Italy in July 2024.
Focus on interdisciplinary research and AI
CAIS researchers will be represented at Easst-4S 2024 in Amsterdam with various contributions on the topics of interdisciplinary collaboration and the narratives and effects of...
Tips for Identifying Fake News and Misinformation for the European Elections
In times when fake news can potentially influence elections, it is crucial for voters to be able to distinguish reliable information from misinformation. This is...
Facts or fakes: How do I recognize trustworthy content on news sites? – Guest: Prof. Dr. Hendrik Heuer
Reliable information is incredibly important for all of us.
Co-Creating Artificial Intelligence: CAIS Launches New Research Program Led by Hendrik Heuer
At the Center for Advanced Internet Studies (CAIS), the new research program “Design of Trustworthy Artificial Intelligence” will start in April under the direction of...
Doctoral and Postdoctoral Researcher: Design of Trustworthy Artificial Intelligence
The Professorship for the Design of Trustworthy Artificial Intelligence at the Center for Advanced Internet Studies (CAIS) gGmbH in Bochum, Germany, offers three positions for Doctoral Researchers and one for a Postdoctoral...
Team
