August 29, 2025
TAIS is a new research programme where the primary objective is to raise awareness of the importance of Trustworthy AI, both academically and publicly. What are we learning so far?
On August 19th – 20th, during a rare sunny week in Vestland county, the official kick-off seminar for the Trustworthy AI Synergy (TAIS) research programme took place in the lovely surroundings of the Solstrand Hotel & Bad.
The seminar was arranged by SLATE, led by SLATE Director Professor Barbara Wasson, the leader of TAIS and the project leader of Trustworthy AI in Education (EduTrust AI). Attending the seminar were Professor Fedor Fomin (Department of Informatics, UiB), the project leader of Algorithmic Foundations of Trustworthy AI, and Valeriya Lyssenko (Department of Clinical Science, UiB), the project leader of TRUSTworthy AI models to predict progression to complications in patients with Diabetes (TRUST-AI4D).
Also attending were professors, researchers, PhD Candidates and Postdoctoral Fellows connected to the three projects.
TAIS facilitates cross-project cooperation, drawing on synergy between the three projects and other research at the University of Bergen (UiB), in Norway and internationally, through various collaboration, innovation and interdisciplinary exchange.
By providing common meeting places, hosting international researchers, supporting interdisciplinary cross-project young researchers networking, organising an international symposium, and promoting joint publications and dissemination, this will lead to more significant and impactful outcomes, particularly in the complex and vital domain of Trustworthy AI.
Day 1: Project Presentations and Ethical Dilemmas
Professor Barbara Wasson gave a presentation of the TAIS programme, aiming towards finding common opportunities and challenges across the three component –projects, each exploring a different aspect of trust in AI systems. The project leaders introduced the three projects that make up the umbrella TAIS programme:
1) the development of algorithms (AlgoTrust),
2) the use of AI in relation to diabetic patients (TRUST-AI4D),
and 3) the use of AI in Education (EduTrust AI).
On the first day, representatives from the Trond Mohn Research Foundation (TMF), who have funded TAIS and the three projects, participated in the seminar.
Associate Professor Simon Knight, an invited Speaker from the University of Technology Sydney (UTS), also participated on the first day. At the end of Day 1, Dr. Knight gave an interesting and inspiring talk entitled “Ethics, Evidence and AI Use: Navigating Interactions”, on the distinction and connection between learning with, and learning about, AI, as well as the ethical dilemmas that arise related to our use of AI. One particularly compelling example covered a parody paper that utilised best practice approaches from the trustworthy literature to develop an algorithm that would mulch elderly people, turning them into a high-nutrient slurry!
Such dilemmas highlight the crucial difference between doing things ethically and doing ethical things. The representatives from TMF contributed to interesting discussions around trust and ethics in the three projects (EduTrust AI, AlgoTrust and TRUST-AI4D), the intersection between AI and humans, and how the three projects connect.
Following the seminar at Solstrand Hotel & Bad, Dr. Knight held another seminar entitled “Democratising AI” on August 21st, as a guest Speaker at SLATE’s offices at UiB.
Day 2: PhDs and PostDocs and a UiB AI Seminar
On the second day, the PhD Candidates and Postdocs associated with the component projects presented their work. These gave interesting insights into research and queries in areas such as the role of trust in AI and society, algorithms and logic, and identification of genes associated with Type 2 diabetic patients.
These very different projects all involve various aspects of trust, from the computational and algorithmic, to the trust of expertise and algorithmic results, right through to layers of trust in complex socio-technical systems.
SLATE Postdoctoral Fellow Anja Salzmann was the first to present on Day 2, with her paper entitled Can we trust technology? Interdisciplinary perspectives on the concept of trust. Dr. Salzmann talked about one of the main challenges that we will face in the TAIS project, namely that the project members represent very different disciplines with different understandings of what trust is— and how to achieve trust and how to research it.
Trust in AI is challenged by a general erosion of public trust in science, trust in experts, trust in the media and political systems that, for example, democracy represents. Dr. Salzmann placed the “trust debate” in a larger academic and historical context and pointed out that trust in AI is not something that can be achieved or made to happen by designing it into technology. It is a relational, very individual phenomenon that is governed by emotions and experiences, and which arises in complex contexts that are difficult if not impossible to manage or control.
Dr. Salzmann underlined the importance of working together to find a common understanding of trust that can create the synergistic effect that one wants to get out of the TAIS collaboration. The presentation sparked a long discussion of the project members' understanding of the word "trust".
The second day ended with group discussions on how we can connect across the various projects to support the TAIS objectives, and the development of plans for disseminating TAIS findings more broadly.
The first public communication activity will take place at the University of Bergen on the 2nd of September 2025, when members of the TAIS project will contribute to the 16th UiB AI seminar, with a focus on Trustworthy AI.
Among the Speakers are Professor Barbara Wasson, SLATE Postdoctoral Fellow Fride Haram Klykken and SLATE-affiliated PhD Candidate Alessia Di Muro (Faculty of Law, UiB). They will be presenting the research project EduTrust AI, which is a collaboration between SLATE, the Faculty of Psychology (UiB), and the Faculty of Law (UiB). Professor Wasson will also be presenting her insights from TAIS and the research programme Trustworthy AI.
