ALCHIMIA results presented in webinar about AI Act and standardization

Oct 2, 2025 | news

On 12 June 2025, Petri Kannisto from BFI presented about the recent results of ALCHIMIA, focusing on the AI Act and standardization! The webinar, called “AI Act and its impact on research, business, and standardization”, was a collaborative effort between four research projects as well as the organizer International Data Spaces Association (IDSA), the moderator being Ilknur Chulani from IDSA.

The webinar started with an introduction to the AI Act, presented by Jan Sturm (Institut Jozef Stefan) from FAME project. The presentation explained, among others, the risk levels in AI systems, the enforcement of the AI Act, and the status and future steps of the regulation.

Next, Sturm continued with another topic, presenting for FAME project. He told hold a solution called StreamStory can provide insights from timeseries with AI and Large Language Models, not forgetting AI explainability. The applied state-of-art data processing methods extract the essential from complex data streams.

Then came Kannisto’s turn to present about ALCHIMIA’s results and its relationship with AI Act and standardization. He explained that ALCHIMIA aims to provide decision support for green steelmaking with an AI-powered, human-factors-aware framework. It seems that AI Act is an important driver for standardization, among other needs in the data-driven landscape. Among the four in-progress standardization contributions from ALCHIMIA, two are related to AI ethics.

The subsequent presentation was about HumAIne project by Pavlos Kranas (LeanXcale). HumAIne develops an operating system for human-AI collaboration in unstructured environments. The project includes considerations for smart manufacturing, smart cities, smart healthcare, smart finance, and smart energy, all of which necessitate considerations for AI ethics.

Next, Sturm presented for the project enRichMyData. The project aims to solve problems related to the increasing complexity in not only AI systems but also the related regulation. The project is very related to AI Act now that it aims to apply Large Language Models for compliance reasoning, “fighting fire with fire”!

After the projects, Sebastion Steinbuss (IDSA) presented about the relationship of International Data Spaces and AI Act. Data Spaces aim for secure, trustable, ownership-aware, discoverable, contract-based data sharing, contributing to interoperability and data-driven ecosystems. Data Spaces are being standardized, for instance, in the draft standard ISO/IEC 20151. The core theme, data governance, is essential to enable human control.

As the penultimate step came a common discussion. It seemed that AI Act is still a new piece of regulation, which means that any organizations working with AI must learn to accumulate knowledge. On the other hand, many AI-related standards, aiming for best practices, are currently not regulatory but rather provides recommendations. In the data-driven side, the Data Space Protocol, providing a standard method for trust and interoperability, will be essential in future AI systems.

Finally, Silvia Castellvi from IDSA wrapped up the session with a landscape of European standardization related to AI, data management, and cloud and edge computing. The participants were encouraged to contribute to the work by joining standardization groups or participating in workshops. In the end, partners from research projects can give valuable insights about missing aspects as well as the needs of industries.

We want to present huge thanks to the organiser and the co-presenters for a successful, insightful webinar! The organizer IDSA is a key driver of interoperability-related standards in the data-driven landscape, driving trust in value chains.

 Please find the webinar recording here: https://www.youtube.com/watch?v=vLQRPjQKIYI

 

Agenda overview:

  • Welcome and goals of the workshop – IDSA 5′
  • Introduction to AI-ACT and regulatory obligations – JSI, Jan Šturm 15′
  • Presentations from the projects on the impact of AI-Act on research, society and business solutions – 50′
    • FAME – Jan Šturm, JSI
    • ALCHIMIA – Petri Kannisto, BFI (VDEh-Betriebsforschungsinstitut)
    • HumAIne – Pavlos Kranas, LeanXcale  
    • enRichMyData – Jan Šturm, JSI 
  • Data spaces for the AI-ACT – IDSA – Sebastian Steinbuss 10′
  • Open discussion (moderated by JSI, Jan Šturm) – 30′: What is the impact of AI-ACT on research and on business, and what the projects are doing about it
  • Wrap up & next steps – IDSA 10′