[Teaching 2026] New academic year and a new place for AI in architectural education

Are we preparing students for the buildings we construct today or for the world in which they will practice tomorrow?

The 2026-2027 materials are ready at the University of Liège: updated books, course booklets, technical details, exercises, readings, and case studies for our architecture and building engineering students. They cover sustainable construction, building components and envelope design, building services and HVAC, circular design in the studio, and building performance simulation and monitoring.

The harder question this year was how generative AI should enter the room. We are already using Tooli in our courses. Many students are using it and benefiting from it. Developed by Buildwise with Embuild, Bouwunie, and Constructiv for the Belgian construction sector, Tooli connects AI assistance with construction practice. Its Nexo assistant helps users find technical information grounded in Buildwise publications and expertise. In our courses, that support comes with an obligation: consult the source, check its applicability, and explain the reasoning.

Thanks to Buildwise, Pieter Van Geite, Tycho de Back, Pepijn Alofs, and Evelyn Meysman, and to our teaching team, Guirec Ruellan, Aurélie Piette, and Charlelie Dagnelie.

In sustainable construction, students first learn how a detail works: structure, water protection, thermal continuity, airtightness, moisture control, materials, and carbon. Only then may they use image generation to communicate a detail they have already validated. When the output defaults to a glass tower in a temperate climate, that bias becomes part of the lesson. A student working on earth construction in the Sahel must question both the image and whose buildings appear in the training data.

In simulation and monitoring, students combine AI with Python, DesignBuilder, and EnergyPlus to draft scripts, modify input files, analyze results, and identify inconsistencies. They remain responsible for verifying their models. Occupant data also requires careful handling: privacy is part of the exercise.

Two questions now sit inside the assignments.

1.What did it cost? Students estimate the electricity and water associated with their AI use, acknowledge the uncertainty, and compare the energy estimate with the operational energy of their building model. Data centers are buildings too.

2.Who is accountable? The person whose name appears on the report.

Under the ULiège Charter for generative AI, our courses require students to disclose tools and prompts, verify outputs, read the assigned papers, and explain their drawings, models, code, assumptions, and results without the machine in the room. AI can support learning. It cannot replace understanding. Students graduating in 2030 may still be designing buildings in 2060. Our job is to build the technical judgment they will need to question and redesign solutions throughout their careers.

The materials are ready. Now the real work begins.

How are you handling AI in your architecture and engineering teaching?

🔗 Course approach: https://www.uliege.be/cms/c_9054334/en/directory?uid=u220815

🔗 ULiège charter: https://www.student.uliege.be/cms/c_19230399/en/uliege-charter-for-the-use-of-generative-artificial-intelligence-in-academic-work

📚 Learn more about our teaching https://www.sbd.uliege.be/cms/c_7654646/en/sbdlab-teaching

🔗 Tooli: https://www.tooli.be

💻 Subscribe to my newsletter https://lnkd.in/diTVT5eq

🅱️ Bilibili b23.tv/bzjL3bn or 🎬 YouTube https://lnkd.in/erHrfkNf

🌐 Explore previous posts and resources https://www.shadyattia.org

🔗 Follow all my professional links, including WeChat 🟩💬 微信https://lnkd.in/eN3xZhhZ

#ArchitectureEducation #BuildingPerformanceSimulation #GenerativeAI ArtificialIntelligence #EngineeringEducation #BuildingTechnology #BuildingScience #GenerativeAI #SustainableConstruction #EmbodiedCarbon #Tooli #Buildwise #BuildingPerformanceSimulation #CriticalThinking #ULiege #SBDLab