Panel on Trust in AI @Digital Life Norway: In AI we trust!? (or don’t we? / should we?)

This October, I had an opportunity to participate in the panel on Trust in AI that took place as part of Digital Life Norway conference organized by Centre for Digital Life Norway (Norwegian University of Science and Technology (NTNU)) that took place in a very peaceful Hurdal (Norway) 🇳🇴🇳🇴🇳🇴.


As part of this panel, together with M. Nicolas Cruz B. (KIWI-biolab), Korbinian Bösl (ELIXIR, both of us being also part of EOSC Association)), and Anamika Chatterjee (Norwegian University of Science and Technology (NTNU)), who masterly chaired this discussion, we discussed trust in AI and data (as an integral part of it), emphasizing the need for transparency, reproducibility, and responsibility in managing them.


What made this discussion to be rather insightful – for ourselves, and, hopefully, for the audience as well – is that each of us represented a distinct stage in the data lifecycle debated upon the aspect of trust and where concerns arise as data moves from the lab to inform AI tools [in biotechnology].
As such we:
✅highlighted the interconnectedness of human actors involved in data production, governance, and application;
✅highlighted the importance of proper documentation to make data usable and trustworthy, along with the need for transparency – not only for data but also for AI in general, incl. explainable AI;
✅discussed how responsibility becomes blurred as AI-driven methodologies become more prevalent, agreeing that responsibility for AI systems must be shared across teams.
Lastly, despite being openness advocate, I used this opportunity to touch on the risks of open data, including the potential for misuse and ethical concerns, esp. when it comes to medical- and biotechnologies-related topics.


All in all, although rather short discussion with some more things we would love to cover but were forced to omit this time, but very lively and insightful. Sounds interesting? Watch the video, incl. keynote by Nico Cruz 👇.

And not of least interest was a diverse set of other events – keynotes, panels, posters etc. – takeaways from which to take back home (not really to home, as from the DNL, I went to the Estonian Open Data Forum, from which to ECAI, and then, finally back home to digest all the insights), where “Storytelling: is controversy good? How to pitch your research to a non-academic audience” by Kam Sripada and panel on supervision are probably the main things I take with me.


Many thanks go organizers for having me and the hospitality, where the later also goes to Hurdal 🇳🇴 in general, as we were lucky enough to have a very sunny weather, which made this very first trip to Norway – and, hopefully, not the last one – very pleasant!

27th European Conference on Artificial Intelligence (ECAI 2024): Celebrating the past, inspiring the future

This October was one of the busiest yet most rewarding months of the year for me. Among several work trips, the highlight was attending the 27th European Conference on Artificial Intelligence (ECAI 2024) in Santiago de Compostela, Spain. Celebrating its 50th anniversary, ECAI remains Europe’s premier venue for AI research and innovation, bringing together thought leaders, researchers, and industry professionals from around the world.

This year’s theme, “Celebrating the Past, Inspiring the Future,” captured the spirit of ECAI’s half-century legacy while driving forward-looking discussions on the next era of artificial intelligence. With over 1,500 participants from 59 countries (so not so very European conference anymore, but rather a global event) and a packed schedule of more than 150 events, among which:

  • “Towards Real-World Fact-Checking with Large Language Models” keynote talk by Iryna Gurevych, (Technische Universität Darmstadt), reflecting on advancements in using language models for verifying information in real time;
  • “Robots (Still) Need Humans in the Loop,” keynote talk by Iolanda Leite (KTH Royal Institute of Technology), who underscored the essential role humans play in AI-driven robotics, even as systems grow more autonomous;
  • “Economic Complexity: Using Machine Learning to Understand Economic Development” keynote talk by Cesar A. Hidalgo (Toulouse School of Economics & Corvinus University of Budapest) that examined how machine learning is transforming our understanding of economic trends and predictions.

These were accompanied with a range of panels, with a few sessions that stood out (my personal opinion though):

  • Economic Impact of AI: Threats and Opportunities with Jeremy Rollison (Microsoft Corporation), David Autor (MIT), and Raquel Jorge Ricart (Elcano Royal Institute) on AI’s potential to reshape labor markets and economies around the world;
  • AI Regulation: The European Scenario (Kilian Gross, Dr. Clara Neppel, IEEE, Beatriz Alvargonzalez Largo, European Commission, José Miguel Bello Villarino, ARC Centre of Excellence for Automated Decision-Making and Society), addressed regulatory considerations;
  • 50th Anniversary Session on the History of AI in Europe paying tribute to AI’s history in Europe, with Luc Steels, Stefano Cerri, Fredrik Heintz, and Tony Cohn sharing reflections on past achievements and a “follow-up” on it in the Future of AI: The Next 50 Years with Fredrik Heintz, Iryna Gurevych, José Hernández-Orallo, Ann Nowe, Toby Walsh;
  • Designing Ethical and Trustworthy AI Research Policies for Horizon Europe centered on ethical standards and trustworthy AI research practices within the EU’s Horizon program, led by Mihalis Kritikos from the European Commission;
  • Funding your Scientific Research with the European Research Council (ERC) with Enrique Alba.

As part of this conference, I had pleasure of presenting a paper co-authored with my former student Jan-Erik Kalmus, based on his Master’s thesis, which I had the privilege of supervising. Our paper, To Accept or Not to Accept? An IRT-TOE Framework to Understand Educators’ Resistance to Generative AI in Higher Education,” examined what barriers might prevent educators from adopting Generative AI tools in their classrooms? Since the public release of ChatGPT, there has been a lively debate about the potential benefits and challenges of integrating Generative AI in educational contexts. While the technology holds promise, it has also sparked concerns, particularly among educators. In the field of information systems, Technology Adoption models are often used to understand factors that encourage or inhibit the use of new technologies. However, many existing models focus primarily on acceptance drivers, often overlooking the unique barriers that educators face. This study seeks to fill that gap by developing a theoretical model specifically tailored to identify the barriers that may prevent educators—academic staff in particular—from integrating Generative AI into their teaching. Our approach builds on Innovation Resistance Theory, augmented by constructs from the Technology-Organization-Environment (TOE) framework. With the designed mixed-method measurement instrument, combining quantitative data with qualitative insights, to capture educators’ specific concerns around Generative AI adoption in higher education, our model has been applied in real-world settings, specifically focusing on Estonian higher education institutions. We examined whether academic staff at public universities in Estonia – often referred to as a “digital nation” – show reluctance toward Generative AI use in educational settings. Preliminary findings highlight several concerns unique to educators, which may shape how Generative AI is integrated into teaching practices.

A Heartfelt Thanks to ECAI’s Organizers – the European Association for Artificial Intelligence (EurAI), the Spanish Artificial Intelligence Society, CiTIUS (Research Centre on Intelligent Technologies), and, of course, the city of Santiago de Compostela for being such a welcoming place.

The Estonian Open Data Forum – Celebrating Progress and Recognizing Achievements

This October, I had the distinct honor of participating in Estonia’s premier event on open data, the Open Data Forum (Avaandmete foorum), organized by the Ministry of Economic Affairs and Communications of Estonia, invited to talk about the role of academia and private sector in the open government data landscape. This annual gathering brings together industry experts, academic researchers, and government leaders to discuss key trends, achievements, and the future of open data in Estonia, along with highlighting the contributions coming from universities awarding the best dissertations developed by Estonian students. The latter made this event very special for me, as one of my students – Kevin Kliimask – was awarded for his outstanding bachelor’s thesis 🏆 🥇 🏅!

In his thesis –“Automated Tagging of Datasets to Improve Data Findability on Open Government Data Portals,” Kevin developed an LLM-powered interface to automate dataset tagging in both English and Estonian, thereby augmenting metadata preparation by data publishers and improving data findability on portals by users – as the practice shows their presence tend to be a challenge. E.g., our analysis conducted on the Estonian Open Data Portal, revealed that 11% datasets have no associated tags, while 26% had only one tag assigned to them, which underscores challenges in data findability and accessibility within the portal, which, according to the recent Open Data Maturity Report, is considered trend-setter. The developed solution was evaluated by users and their feedback was collected to define an agenda for future prototype improvements. The thesis has been already transformed into the scientific paper 👉 TAGIFY: LLM-powered Tagging Interface for Improved Data Findability on OGD portals presented at the IEEE international conference (I posted on this earlier 👉 here).

As for my talk titled Unlocking the Power of Open Data: The Role of Academia and the Private Sector in Building Inclusive and Sustainable Open Data Ecosystems, I emphasized the need for a holistic approach to open data that transcends merely opening/publishing data, rather requiring adopting a systemic view that considers an open data initiative as an Open Data Ecosystem (also confirmed by Open Data Charter 👉here), as we deal not only with open data (availability), portal, stakeholders, actors, but also processes surrounding them, emerging technologies & different forms of intelligence, going beyond just Artificial Intelligence, whose role, however, is crucial (see our paper on the eight-fold role of AI in OGD).

As such, while discussing the main idea of ​​the talk – the role of academia & the private sector in the ODE, which as per me is at least four-fold, namely – data consumers, data providers, contributors to ODE sustainability & myth busters on the global stage (assigning “Made in Estonia” tag to OGD in addition to the one we have for e-government), I also expanded the general mantra of “Data For AI” to “data for AI, AI for data, data not only for AI and not only AI for data”.


A big thank you to the organisers, who gathered so many speakers (Cybernetica, FinEst Centre for Smart Cities, Riigi Infosüsteemi Amet // Estonian Information System Authority (NCSC-EE), University of Tartu and many others) to discuss the highlights of today & tomorrow for Estonian Open Data and High-Value Datasets in particular as it was the main focus of the Forum – was happy to be part of these discussions!

The University of Tartu ranks among the top 1% the world’s most highly-cited research institutions & Stanford Elsevier Top Scientists List

Did you know that the University of Tartu ranks among the top 1% the world’s most highly-cited research institutions in the world? This [but not only, of course] makes it one of Northern Europe’s leading universities and the highest-ranked university in the Baltics (according to Times Higher Education). Not bad, especially considering that Estonia is the smallest Baltic country in both area and population & the third most sparsely populated in the EU, isn’t it?

This is also somewhat supported by the latest Stanford Elsevier Top Scientists List of what they call to be “the top 2% of scientists“, which, as any list of this kind, should not be taken too seriously, but rather with an extreme caution, since of course quantitative metrics such as the number of publications and citations are far from perfect indicators for measuring research quality and impact❗ But, in the lack of alternatives…

Indeed all the “be cautious” for various reasons* and “do not overestimate those who is in it” that have been discussed at length should be “on,” going through the list, with reasons spanning from serious cases of scientific misconduct to the fact that, according to the list, Einstein published for over 120 years, with his recent paper in 2021 (perhaps we should stop complaining on our own schedules :D) and several other debates around the list (see here, here & here). However, nonetheless, finding yourself among colleagues you truly respect does make the colors around a bit brighter (though in the single recent year impact category (again, citations-based “impact”)).

Back to the point, more than 40 researchers from the University of Tartu—including 4 from my own University of Tartu Institute of Computer Science, incl. myself (? not sure of the total # as there are 300+ people in the Institute, so the count is filtered by Artificial Intelligence, ICT, Information Systems)—listed in Stanford’s single-year category, with slightly below 40 – career-long!

As for me, I’m taking this as a green light to treat myself by grabbing some yummy that I feel somewhat deserved 🍨 🚀as this is a sort of – one small step for research and the university (or not so small), one giant leap for an early career researcher (in the age of 18 years, with 10+ years of experience :D)

CFP for Special Issue in IEEE Transactions on Technology & Society: Trustworthy Data Ecosystems for Digital Societies

IEEE Transactions on Technology & Society launches the new Special Issue on the “Trustworthy Data Ecosystems for Digital Societies“, edited by Asif Gill, Anastasija Nikiforova, Ina M. Sebastian, Martin Lnenicka, Anushri Gupta. On behalf of the editors of this SI, I sincerely invite you to consider submitting your work to it.

Key topics surround intersection of data ecosystem and AI topics, i.e., AI in and for trustworthy data ecosystems, and include, but are not limited to:

  • Impact of trustworthy data ecosystem on digital societies at the local, national and global levels
  • Conceptualization of trustworthy data ecosystems domains and characteristics for digital societies
  • Data trust regulations, polices, strategies and standards
  • Trustworthy data ecosystem infrastructure as a social construct
  • Trustworthy data ecosystem architecture, interfaces, methodologies, orchestration, patterns, solutions, and technology platforms
  • System and data quality, governance, security, privacy, protection, and safety
  • Data linking, interoperability, sharing and observability

Read more in the below CFP or here.