Mohamed bin Zayed University of Artificial Intelligence

How can AI research help shape a path in aerospace engineering? For Huanying “Joy” Yeh, it landed her at SpaceX, where she’s an electrical engineer.

A member of MBZUAI’s Undergraduate Research Internship Program (UGRIP) 2024 cohort, she explored how LLMs can be prompted to solve Olympiad linguistics problems, developing the problem-solving and systems thinking that translates across engineering disciplines, while building skills in project management and cross-team collaboration.

Through UGRIP, Joy saw how AI intersects with aerospace, computer science, and complex engineering systems.

Stay tuned for more stories from our past UGRIP participants and updates from the 2026 cohort.

Learn more: https://mbzuai.ac.ae/study/

At CVPR 2026, MBZUAI researchers will present work addressing core challenges in computer vision and multimodal AI: how models reason over long videos, how systems generalise with limited labels, how AI agents can support remote sensing, and how the field should approach watermarking in generative media.

From the many papers accepted, we're spotlighting a selection of oral and highlight papers which focus on visual reasoning, trustworthy media, domain adaptation, remote sensing, and efficient foundation-model methods.

Genetics explains only a fraction of why people get sick. Environment, lifestyle, and the epigenome explain far more. And unlike DNA, those factors can change.
A new framework published in Nature Genetics sets out how the UAE is uniquely positioned to turn that insight into population-scale disease prevention.

The paper was co-authored by MBZUAI's:
Professor Eran Segal - Dean of the Biological and Life Sciences Division, and Professor of Computational Biology;
Professor Eduardo Beltrame - Assistant Professor of Computational Biology;
Professor Shahrukh Hashmi- Adjunct Professor of Personalized Medicine, and Director of Research, Department of Health Abu Dhabi (DoH);
Professor Andrew Feinberg - Visiting Scholar, MBZUAI, Director of the Center for Epigenetics, Chief of the Division of Molecular Medicine in the Department of Medicine, and the King Fahd Professor of Medicine, Oncology, Molecular Biology, and Genetics in the School of Medicine at Johns Hopkins University.

Read more 👉 https://mbzuai.ac.ae/news/how-population-scale-epigenetics-could-transform-healthcare-in-the-uae/

Link to paper here 👉 https://www.nature.com/articles/s41588-026-02630-x.epdf?sharing_token=5R1MYee2S7tlE8VX2-dlU9RgN0jAjWel9jnR3ZoTv0PVd6ngVdcJ7rYjlIc6xgP7_RMsBCZZ_cyTPDJFkvgpe7bGT9Op9SdIsDe1GCJsFLTJK7h7wRvi8JUu5C0KBq4qyRbF94KFoy-wb-bLjvavF5umMGCOkt_zlQBeTN36YE0%3D

تتمنى جامعة محمد بن زايد للذكاء الاصطناعي لكم ولأحبائكم عيد أضحى مبارك. كل عام وأنتم بخير!✨

MBZUAI wishes you and your loved ones a blessed Eid Al-Adha. ✨

What if language models are not just text generators, but cultural archives?

New research from MBZUAI and the Technical University of Darmstadt finds that large language models encode cultural commonsense: not just facts, but the chains of actions and expectations that shape everyday life across different societies.

The catch? Models tend to express that knowledge more reliably in English than in the native language of the culture itself.

The team tested their approach across five countries: China, Indonesia, Japan, England, and Egypt. They found that adding structured cultural knowledge helped smaller models perform better on cultural reasoning tasks.

It is not a complete fix, but it is a step toward AI systems that can navigate different social contexts without flattening them.

Research by Junior Cedric Tonga, Chen Cecilia Liu, Iryna Gurevych, and Fajri Koto.

Read more: https://mbzuai.ac.ae/news/ai-models-are-becoming-cultural-archives/

Richard Morton, Vice President and Executive Director of the Institute of Foundation Models, took the stage at the Abu Dhabi AI Native Summit 1.0 this week.

Highlights from his talk include:

1. The UAE isn’t just deploying AI — it’s building frontier models from scratch. MBZUAI sits alongside leading global tech organizations capable of this.
2. Sovereignty is the overlooked variable. For governments embedding AI into core processes, open, transparent models aren’t a nice-to-have; they’re a must-do. K2 and JAIS are fully open-source and UAE-built for exactly this reason.
3. Agentic AI is happening now. End-to-end procurement, unified citizen services, executive workflows are all possible without human intervention. The challenge isn’t the technology; it’s knowing where to pilot it first.
4. And world models — simulating physical reality for robotics and autonomous systems — are the next big frontier, with MBZUAI among only three entities globally working at this level.

Exciting things ahead for and from the UAE. 🇦🇪

For more info, visit: https://bit.ly/4nIQxMQ

استقبلت جامعة محمد بن زايد للذكاء الاصطناعي دفعة جديدة في برنامج الماجستير في الذكاء الاصطناعي التطبيقي من خلال بعثات محمد بن راشد الحكومية.

يُعد البرنامج الأول من نوعه في دولة الإمارات، ويهدف إلى تزويد الكوادر الحكومية الإماراتية بمهارات متقدمة في مجال الذكاء الاصطناعي، لتسريع تبنيه في الجهات الحكومية ودعم استراتيجية الإمارات للذكاء الاصطناعي 2031.

يشمل البرنامج مجموعة من المحاور التخصصية مثل الذكاء الاصطناعي التوليدي والتعلم العميق، بالإضافة إلى الروبوتات والذكاء الاصطناعي الموثوق وأنظمة المدن الذكية. يجمع بين التميز الأكاديمي والتطبيق العملي، مما يهيئ خريجيه لقيادة التحول في استخدام الذكاء الاصطناعي في القطاع الحكومي بدولة الإمارات.

MBZUAI welcomes a new cohort to the Master of Applied Artificial Intelligence (MAAI) program through the Mohammed bin Rashid Government Scholarships.

This UAE-first program will equip Emirati government professionals with advanced, practical AI skills to accelerate AI adoption across government entities and support the UAE Artificial Intelligence Strategy 2031.

From generative AI and applied deep learning to robotics, trustworthy AI, and smart city systems, the program combines academic excellence with real-world application, preparing graduates to lead AI transformation across the UAE public sector.

ae