UZH.ai

UZH.ai is proud to sponsor SwissText 2026, Switzerland’s conference for text analytics, NLP, and language AI. Organized by the Swiss Association for Natural Language Processing (SwissNLP), the 11th annual SwissText conference brings together text analytics experts from industry and academia to exchange ideas, explore real-world applications, and discuss the latest developments in language technologies. Join us at the University of Zurich on June 10, 2026, and be part of the conversation shaping the future of text and AI. Sign up here: https://lnkd.in/d46TqvSp #SwissText2026 #UZHai #SwissNLP #NLP #TextAnalytics #LanguageAI #AI

What does ethical AI really mean? In a new Swiss Academy of Engineering Sciences SATW Technology Outlook interview, Dr Ning Wang from University of Zurich explains why ethics cannot be added to AI at the end. The discussion focuses on Apertus, the open large language model developed by ETH Zürich and EPFL through the Swiss AI Initiative. Apertus makes its code, training data and methods openly available - a big step for trustworthy AI in Switzerland and Europe. Dr. Wang highlights why ethical thinking must be part of AI development from the very start - not something added only at the end. While transparency is an important first step, AI also shapes whose knowledge is seen, which languages are included, and which voices may be left out. If ethics comes too late, bias can become built into the system - and much harder to fix. Read the full interview: https://lnkd.in/egGDn_NN #ResponsibleAI #AIEthics #SwissAI #Apertus #UZHai

Exciting new paper on AI-enabled drone flight in confined spaces! In “Low-Latency Event-Based Velocimetry for Quadrotor Control in a Narrow Pipe,” Prof. Davide Scaramuzza and Leonard Bauersfeld from the Robotics and Perception Group at the University of Zurich show how AI can help drones fly more safely in one of the most challenging environments: narrow pipes. The work combines event-based vision with AI-based control, enabling a quadrotor to detect airflow disturbances in real time and respond before they push it toward the pipe wall. Instead of relying only on pre-programmed reactions, the drone uses learning-based methods to interpret fast-changing visual signals, estimate the invisible forces around it, and adjust its flight to stay stable. A great example of how AI, robotics, and perception come together to push autonomous systems into places that are difficult or dangerous for humans to access. Read the full paper: https://lnkd.in/eVwxsNcs Video: https://lnkd.in/eqAHKM8h #AutonomousSystems #Drones #EventBasedVision #MachineLearning #Robotics

At this year’s Dies academicus, AI took the stage in a very UZH way: curious, playful and responsible. For the University of Zurich's 193rd anniversary, President Michael Schaepman spoke with an AI-generated twin of himself. What looked like a fun experiment also carried an important message: AI is changing teaching, learning and research, and universities must help shape this future responsibly. The event also highlighted real AI impact in research. Congratulations to Prof. Florian A. Wenzl, recipient of the Walter and Gertrud Siegenthaler Foundation Science Award. His research combines AI with new cardiovascular biomarkers to better predict risk in acute coronary syndromes, including heart attacks Read more about the Dies academicus: https://lnkd.in/e3u9ZACc #ResponsibleAI #ArtificialIntelligence #Research #HigherEducation #UZH

Exciting to be part of a region where AI research, innovation, and real-world applications come together so strongly. Greater Zurich Area

How can AI help researchers move from literature discovery to scientific thinking? PaperScope is a digital workspace, built by University of Zurich researcher Ushnik Das, that helps researchers discover relevant papers, organise ideas, and map connections across the literature. The platform combines personalised paper discovery with a visual research canvas, where users can connect papers, concepts, hypotheses, and questions. It also includes weekly paper digests, AI-guided discovery, citation network exploration, and support for multiple interest profiles. More than 500 researchers have joined PaperScope, with early adoption across UZH, ETH, and beyond. The platform is being developed as a practical tool for discovery, organisation, and visual reasoning in research. Explore PaperScope: https://lnkd.in/e-BNCQQg #AIForScience #Research

How can we make AI systems easier to understand, more transparent and better aligned with human needs? Join the next IfI Colloquium for a talk by Prof. Dr. Matthias Zeppelzauer from USTP – University of Applied Sciences St. Pölten, hosted by Prof. Dr. Juergen Bernard from the UZH Department of Informatics. 💡 The session will focus on human-centric machine learning and explore how transparency, explainability and human feedback can help shape AI systems that are more useful, fair and trustworthy. 📍 Location: BIN 2.A.01, Department of Informatics, Binzmühlestrasse 14, Zurich 📅 Date: 7 May 2026 🕔 Time: 17:15–18:30 No registration required. More information: #ArtificialIntelligence #MachineLearning #ExplainableAI

What does it mean to shape AI responsibly at a university? The UZH Annual Report 2025 offers a strong answer: by bringing people, disciplines and ideas together. Across the University of Zurich, AI is becoming part of how we research, teach, learn and work. For UZH.ai, this momentum reflects exactly why the hub exists: to connect expertise across faculties, support responsible innovation, and create a space where AI is explored not only as a technology, but as a societal, scientific and human question. In his Annual Report interview, UZH President Michael Schaepman asks “what kind of future we want.” He also reminds us that “diversity is the key” to UZH’s future-readiness. That is also the promise of AI at UZH: diverse perspectives, shared knowledge and a commitment to shaping AI in ways that serve people, science and society. Read the full interview with UZH President Michael Schaepman and Director of Education Silvia Steiner: https://lnkd.in/eH3S4H-8 #AI #ResponsibleAI #DigitalTransformation #Research #Education #Innovation

Does Human-centred AI need a human-centred dialogue? We’re pleased to share the upcoming expert panel “CITL: The Clinician in the Loop AI and Clinical Uncertainty.” Clinical AI deployments frequently stall because predictive uncertainty isn't communicated clearly, and escalation pathways for clinician oversight remain undefined. The Clinician in the Loop AI (CITL) project at the UZH Digital Society Initiative is addressing this challenge directly. Bringing together experts from medicine, engineering, law, and ethics, the panel will explore how uncertainty should be communicated in AI-assisted care, where human intervention is non-negotiable, and how the CITL Playbook can help define actionable best practices. On May 8, 11:00–12:00, the discussion will examine how clinician-in-the-loop AI can be built for transparency, oversight, and real clinical use. The panel will be followed by lunch and networking. Seats are limited. Register here: https://lnkd.in/eWMaKpP3 #UZHAI #ClinicalAI #HumanCentredAI #AIinHealthcare #DigitalHealth #CITL

"It’s here now. It’s real!" Physical AI is leaving the lab and entering our cities, workplaces, and public spaces – and from 1–4 July 2026, Davos will become a “Robot City” to explore this shift up close. As UZH.ai, we are partnering with Davos Tech Summit because we see a unique opportunity – and responsibility. Our role is not only to advance robotics and AI technologies, but to ask the critical questions: – How do robots and intelligent machines interact with people in everyday environments? – What does this mean for safety, trust, inclusion, and democracy? – How do we design governance, law, and ethics around physical AI that truly serve society? As the University of Zurich’s hub for AI, we bring together researchers from computer science, robotics, medicine, law, humanities, and social sciences to examine these questions from multiple perspectives and in dialogue with industry, policymakers, and the public. Davos Tech Summit – “Touching Intelligence” – offers a real‑world testbed: a robot city across Davos, four tracks from technology to responsibility and human factor, and a congress where global experts will explore how advanced robots will act in and interact with the physical world. We are excited to contribute to this conversation and to help shape what responsible physical AI should look like – in Davos and beyond. #UZHai #physicalAI #DavosTechSummit #RobotCity #ResponsibleAI Ruben Kranendonk Pascal Kaufmann (Mindfire, Lab42), Gion Sialm (University of Applied Sciences of the Grisons), Kristof A. Hertig (digitalswitzerland), Eric Anderegg (GenAI Zürich), Jan Kerschgens (Swiss National Science Foundation SNSF), Hanna Brahme (AI House Davos), Daniel Naeff (ETH AI Center), Rebecca Brauchli (ZHAW digital)