CMU Human-Computer Interaction Institute

Congratulations to the Class of 2026 and best wishes to the nearly 100 graduate students who earned their master's degrees with us over the course of the year.

Join us in celebrating the students (now ) from these 2 graduate programs:
🎓 Master of Human-Computer Interaction (MHCI)
🎓 Master of Educational Technology & Applied Learning Science (METALS)

We'll see some of you back in the classroom this week for a strong finish on those Capstone projects!

📸: Huge thanks to Lorraine and Jenn, academic program managers, for the photos.
Carnegie Mellon University School of Computer Science

We're proud to share 9 more papers from . These papers with Carnegie Mellon University contributing authors received Honorable Mention awards, a distinction for the top 5% of papers. Congratulations, authors and teams!

Learn more about CMU's work coding greener websites, personalizing mindfulness activities, sharpening eye tracking accuracy, accelerating hands-on trade skills, and more!

🌱💻 EcoAssist: Integrating energy feedback into AI-assisted coding to help developers build greener websites without losing speed.

🔎🩺 Evidotes: A browser extension that augments peer health posts with scientific evidence and personal experience to support chronic illness patients through user-selectable lenses.

👁️🖥️ HiFiGaze: A new approach that uses screen reflections in the eye to significantly improve eye-tracking accuracy on consumer devices.

🧠🤝 LLMs in Behavioral Health Services: Exploring how AI can scale and support peer-run mental health services while preserving the trust and lived experience at the core of care.

🧘📲 MindfulAgents: Using an expert-aligned multi-agent system to deliver personalized mindfulness meditation and improve user engagement and mindfulness outcomes over time.

🔧🔒 Privy: A tool that helps non-privacy experts create high-quality privacy impact assessments for early-stage AI products.

🧜‍♀️💬 The Siren Song of LLMs: LLM Dark Patterns investigates how large language models may subtly steer, pressure, manipulate or mislead users through interaction design and conversational framing.

👩‍🏭🥽 WELDAR: Moving beyond simulation, WeldAR shows that in-situ augmented reality (AR) guidance during real-world practice can effectively support the learning and transfer of complex manual skills training.

🧗💡 Will They Try Again?: A large-scale EdTech field study of 164,000+ students and 17 million problems shows how different persuasive design strategies helped students persist after failure, and their effects added together without redundancy.

We would love to tell you more about all of these, but we're at the character limit! So, please visit our website where you can learn more and read these papers.
🔗 Details: https://hcii.cmu.edu/news/cmu-chi-2026

Three papers with CMU and HCII authors received Best Paper awards at CHI 2026. Learn more about making walls that move, automating accessible PDF metadata, and writing sticky stories for responsible AI work.

🤖🏛️ Towards Fluent Interaction with Cyber-Physical Architecture

What happens when your walls begin to move? CMU’s latest research on robotic environments reveals that preserving user autonomy in automated spaces requires new, modality-agnostic models of human intent.
Authors: + + +
————————

📝✅ iTagPDF: Towards Finally Automating PDF Accessibility

iTagPDF automatically generates accessibility metadata (tags, reading order, and alt text) for academic PDFs by jointly modeling the PDF and its source.
Authors: +
————————

🍯📖 “I Don’t Think RAI Applies to My Model” – Engaging Non-champions with Sticky Stories for Responsible AI Work

We show how “sticky” AI-generated stories can help practitioners who think responsible AI doesn’t apply to them actually see why it does.
Authors: + CMU S3D
————————

Learn more on our website: https://hcii.cmu.edu/news/cmu-chi-2026