The first Tertulia AI of 2026 placed Responsible AI centre stage; researchers at INESC TEC have built a community in constant interaction and, through this event, showcase different perspectives on the topics under discussion.
The second Tertulia AI dived into the “deep waters” of Artificial Intelligence and, without losing footing, presented concrete examples of the work being carried out “in-house”. It provided a platform for the many dimensions surrounding Responsible AI. To guide the debate, INESC TEC researchers took to the stage to explain how they are using AI in a responsible, transparent and robust way, and to explore how they can connect the pieces and collectively address future challenges.
These sessions reflect the INESC TEC community’s ambition to create a space for collaboration and the exchange of ideas at a time when Artificial Intelligence has been propelled beyond research centres and into newspaper headlines and everyday conversations about how it may shape the future.
The opening minutes of the session set out the “state of the art”, with a review of several industry developments from recent months. Eight researchers (most of them from INESC TEC) moved to demonstrated how they are applying AI responsibly. Topics ranged from the technological components required for real-world use cases to multidimensional model quality metrics, from code quality and the importance of data quality to the urgent need for a catalogue with model IDs that captures how systems position themselves across the various dimensions of Responsible AI.
From this “melting pot of ideas” came several key questions: is it possible to combine all individual contributions into a single, coherent perspective and help shape the European agenda? Is the institution equipped to develop Responsible AI that is useful both internally and externally? The answers will be developed internally under a “better together” approach, explained Ana Costa e Silva, researcher at INESC TEC and one of the initiative’s coordinators.
“From this edition, we have drawn several conclusions and a sense of optimism. By combining our individual contributions, we can create a more robust environment, not only for ourselves but also for the companies that collaborate with us in the future,” she mentioned.
Always connected
Before exploring Responsible AI, the first edition of the Tertulia AI had already invited the community to reflect on the “breakneck pace” of AI’s evolution.
Last July, discussions covered developments in machine learning and deep learning, Agentic AI, real-time Edge AI, multimodal AI, foundation models, the relationship between AI and robotics, synthetic data simulation, and the paths being followed by industry and academia – including who is leading the way.The theme of the next edition will emerge from community discussion and may focus on AI applied to robotics or on Agentic AI.
The topic is always chosen by participants, as the event is simply the most visible expression of an ongoing, everyday discussion at INESC TEC: “We have created an online community to bring people closer together, foster synergies and strengthen a sense of sharing and belonging. It is a space where we can talk with colleagues working in AI, read and share news, disseminate papers, launch polls, find partners, ask questions and collaborate – a space built by all of us.”
News, current topics, curiosities and so much more about INESC TEC and its community!
