General AI research publications from MBZUAI, KAUST, TII, QCRI, and other GCC institutions, including conference papers, journal articles, and preprints.
KAUST and SARsatX have developed a method using Generative Adversarial Networks (GANs) to generate synthetic SAR imagery for training deep learning models to detect oil spills. Starting with just 17 real SAR images, they generated over 2,000 synthetic images to train a Multi-Attention Network (MANet) model. The MANet model, trained exclusively on synthetic data, achieved 75% accuracy in identifying oil spill areas, matching the performance of models trained on larger real datasets. Why it matters: This advancement enables faster and more reliable environmental monitoring using AI, even when real-world data is scarce, reducing the need to wait for actual disasters to occur.
The paper introduces the Prism Hypothesis, which posits a correspondence between an encoder's feature spectrum and its functional role, with semantic encoders capturing low-frequency components and pixel encoders retaining high-frequency information. Based on this, the authors propose Unified Autoencoding (UAE), a model that harmonizes semantic structure and pixel details using a frequency-band modulator. Experiments on ImageNet and MS-COCO demonstrate that UAE effectively unifies semantic abstraction and pixel-level fidelity, achieving state-of-the-art performance.
The paper introduces AraToken, an Arabic-optimized tokenizer based on the SentencePiece Unigram algorithm that incorporates a normalization pipeline to handle Arabic-specific orthographic variations. Experiments show that AraToken achieves 18% lower fertility compared to unnormalized baselines. The Language Extension Pipeline (LEP) is introduced to integrate AraToken into Qwen3-0.6B, reducing evaluation loss from 8.28 to 2.43 within 800 training steps on 100K Arabic samples. Why it matters: This research provides an efficient tokenizer tailored for Arabic, improving performance of LLMs on Arabic text and benefiting Arabic NLP research by providing released resources.
A new benchmark, LongShOTBench, is introduced for evaluating multimodal reasoning and tool use in long videos, featuring open-ended questions and diagnostic rubrics. The benchmark addresses the limitations of existing datasets by combining temporal length and multimodal richness, using human-validated samples. LongShOTAgent, an agentic system, is also presented for analyzing long videos, with both the benchmark and agent demonstrating the challenges faced by state-of-the-art MLLMs.
KAUST and the International Maize and Wheat Improvement Center (CIMMYT) have signed an MoU to collaborate on developing climate-resilient food crops. The collaboration will combine CIMMYT’s expertise in maize and wheat breeding with KAUST’s strengths in genomics and computational agriculture. The partnership will focus on genomic selection, data analytics, and digital breeding technologies, including capacity-building programs. Why it matters: The partnership aims to enhance food security in Saudi Arabia and the wider region by developing resilient, high-yielding crop varieties suited to arid environments.
The paper introduces OmniGen, a unified framework for generating aligned multimodal sensor data for autonomous driving using a shared Bird's Eye View (BEV) space. It uses a novel generalizable multimodal reconstruction method (UAE) to jointly decode LiDAR and multi-view camera data through volume rendering. The framework incorporates a Diffusion Transformer (DiT) with a ControlNet branch to enable controllable multimodal sensor generation, demonstrating good performance and multimodal consistency.
A new study compares vision-language models (VLMs) to YOLOv8 for wastewater treatment plant (WWTP) identification in satellite imagery across the MENA region. VLMs like Gemma-3 demonstrate superior zero-shot performance compared to YOLOv8, trained on a dataset of 83,566 satellite images from Egypt, Saudi Arabia, and UAE. The research suggests VLMs offer a scalable, annotation-free alternative for remote sensing of WWTPs.
The UK-Qatar Joint AI Research Commission, involving institutions such as Queen Mary University of London, has published its final report. This document consolidates findings and recommendations from collaborative research efforts between the United Kingdom and Qatar focused on artificial intelligence. The report signifies the conclusion of a structured initiative aimed at advancing AI knowledge and fostering scientific cooperation between the two nations. Why it matters: This collaboration strengthens international research ties and provides strategic insights that could influence future AI development and policy in Qatar and the wider Middle East.
KAUST researchers discovered that the red algae strain Galdieria yellowstonesis can convert sugars from chocolate-processing waste into C-phycocyanin, a valuable blue pigment. The study found that high levels of carbon dioxide promote Galdieria growth, and the resulting phycocyanin was deemed food-safe by the U.S. FDA. Mars supported the research by providing chocolate samples. Why it matters: This research offers a sustainable method for waste management and contributes to a circular economy in the region, with potential applications in food, cosmetics, and pharmaceuticals.
Researchers at MBZUAI introduce "Interactive Video Reasoning," a new paradigm enabling models to actively "think with videos" by performing iterative visual actions to gather and refine evidence. They developed Video CoM, which reasons through a Chain of Manipulations (CoM), and constructed Video CoM Instruct, an 18K instruction tuning dataset for multi-step manipulation reasoning. The model is further optimized via reinforcement learning with reasoning aware Group Relative Policy Optimization (GRPO), achieving strong results across nine video reasoning benchmarks.
Researchers at MBZUAI have introduced Video-R2, a reinforcement learning approach to improve the consistency and visual grounding of reasoning in multimodal language models. Video-R2 combines timestamp-aware supervised fine-tuning with Group Relative Policy Optimization (GRPO) guided by a Temporal Alignment Reward (TAR). The model demonstrates higher Think Answer Consistency (TAC), Video Attention Score (VAS), and accuracy across multiple benchmarks, showing improved temporal alignment and reasoning coherence for video understanding.
The paper introduces FanarGuard, a bilingual moderation filter for Arabic and English language models that considers both safety and cultural alignment. A dataset of 468K prompt-response pairs was created and scored by LLM judges on harmlessness and cultural awareness to train the filter. The first benchmark targeting Arabic cultural contexts was developed to evaluate cultural alignment. Why it matters: FanarGuard advances context-sensitive AI safeguards by integrating cultural awareness into content moderation, addressing a critical gap in current alignment techniques.
This paper analyzes the energy consumption and carbon footprint of LLM inference in the UAE compared to Iceland, Germany, and the USA. The study uses DeepSeek Coder 1.3B and the HumanEval dataset to evaluate code generation. It provides a comparative analysis of geographical trade-offs for climate-aware AI deployment, specifically addressing the challenges and potential of datacenters in desert regions.
Luma AI, a startup developing multimodal AGI, has raised $900 million in a funding round led by HUMAIN, a Saudi Arabia-based investment firm. The company is focused on building general-purpose AI models that can understand and generate different types of data, including images, video, and 3D scenes. This funding round will allow Luma AI to scale its research and development efforts. Why it matters: This investment signals the growing interest and financial commitment from Saudi Arabian entities in advancing artificial general intelligence capabilities.
Researchers at MBZUAI have introduced EvoLMM, a self-evolving framework for large multimodal models that enhances reasoning capabilities without human-annotated data or reward distillation. EvoLMM uses two cooperative agents, a Proposer and a Solver, which generate image-grounded questions and solve them through internal consistency, using a continuous self-rewarding process. Evaluations using Qwen2.5-VL as the base model showed performance gains of up to 3% on multimodal math-reasoning benchmarks like ChartQA, MathVista, and MathVision using only raw training images.
This paper proposes a framework for understanding AI sovereignty as a balance between autonomy and interdependence, considering global data, supply chains, and standards. It introduces a planner's model with policy heuristics for equalizing marginal returns across sovereignty pillars and setting openness. The model is applied to India and the Middle East (Saudi Arabia and UAE), finding that managed interdependence, rather than isolation, is key for AI sovereignty.
Researchers from MBZUAI have developed MMRINet, a Mamba-based neural network for efficient brain tumor segmentation in MRI scans. The model uses Dual-Path Feature Refinement and Progressive Feature Aggregation to achieve high accuracy with only 2.5M parameters, making it suitable for low-resource clinical environments. MMRINet achieves a Dice score of 0.752 and HD95 of 12.23 on the BraTS-Lighthouse SSA 2025 benchmark.
KAUST startup Polymeron is converting date and poultry waste into biodegradable plastics, aligning with Saudi Vision 2030's goals for economic diversification and sustainability. The company aims to process 20,000 tons of waste annually, contributing to a circular economy and creating jobs in polymers, manufacturing, and sales. Polymeron is building its pilot plant at KAUST with support from the National Technology Development Program (NTDP). Why it matters: This initiative addresses plastic pollution and agricultural waste while creating a new industry in the Kingdom, tapping into a global market for biodegradable plastics valued at $13 billion in 2024.
KAUST has 16 researchers on the 2025 Clarivate Highly Cited Researchers list, representing 59% of the total from Saudi Arabia. The list recognizes researchers whose publications rank in the top 1% of citations between 2014 and 2024. Some notable research includes work on land degradation, the Red Sea's blue economy, and quantum random number generation. Why it matters: This recognition highlights KAUST's growing research influence and its ability to attract top international talent, contributing to Saudi Arabia's scientific advancement.
Abu Dhabi is spearheading the development of a comprehensive regulatory framework for air taxis within the UAE. This initiative aims to establish a global model for rules governing urban air mobility. The framework focuses on safely integrating advanced aerial technologies into the future transportation ecosystem. Why it matters: This positions the UAE as a forward-thinking leader in shaping the future of advanced aerial transport and setting international precedents for this emerging industry.
This paper introduces Cross-Document Topic-Aligned (CDTA) chunking to address knowledge fragmentation in Retrieval-Augmented Generation (RAG) systems. CDTA identifies topics across documents, maps segments to topics, and synthesizes them into unified chunks. Experiments on HotpotQA and UAE legal texts show that CDTA improves faithfulness and citation accuracy compared to existing chunking methods, especially for complex queries requiring multi-hop reasoning.
This study compares AI uptake in the UAE and Kuwait, analyzing how constitutional, collective-choice, and operational rules shape AI implementation and its impact on citizen centricity and public value creation. It finds that the UAE's concentrated authority and pro-innovation environment enable scaling AI initiatives, while Kuwait's dispersed governance and cautious approach limit progress despite similar resources. The research highlights the importance of vertical rule coherence over wealth in determining AI's public-value yield.
A new method is proposed to reduce the verbosity of LLMs in step-by-step reasoning by retaining moderately easy problems during Reinforcement Learning with Verifiable Rewards (RLVR) training. This approach acts as an implicit length regularizer, preventing the model from excessively increasing output length on harder problems. Experiments using Qwen3-4B-Thinking-2507 show the model achieves baseline accuracy with nearly twice shorter solutions.
This paper introduces a framework that combines machine learning for multi-class attack detection in IoT/IIoT networks with large language models (LLMs) for attack behavior analysis and mitigation suggestion. The framework uses role-play prompt engineering with RAG to guide LLMs like ChatGPT-o3 and DeepSeek-R1, and introduces new evaluation metrics for quantitative assessment. Experiments using Edge-IIoTset and CICIoT2023 datasets showed Random Forest as the best detection model and ChatGPT-o3 outperforming DeepSeek-R1 in attack analysis and mitigation.
An article published in Nature examines the profound implications of Artificial Intelligence for the workforce within the Gulf Cooperation Council (GCC) countries. It likely delves into the challenges of job displacement and the critical need for reskilling initiatives, alongside the opportunities presented by AI integration across various sectors. The analysis presumably outlines strategies for GCC nations to adapt their labor markets and education systems to the evolving future of work. Why it matters: Understanding these dynamics is crucial for GCC nations to develop effective policies for workforce development, economic resilience, and sustainable growth in an AI-driven era.
Cisco and KAUST have launched a new AI Institute located at the KAUST campus in Saudi Arabia. The institute will focus on AI research, development, and education, including AI-native communication systems and edge infrastructure for Industry 5.0. Cisco will donate its latest technology, including a Cisco AI POD. Why it matters: This initiative will foster AI innovation and develop a skilled AI workforce within the Kingdom, aligning with Saudi Vision 2030 and the nation's digital transformation goals.
Researchers have developed a CNN-based deep learning model for predicting coastal flooding in cities under various sea-level rise scenarios. The model utilizes a vision-based, low-resource DL framework and is trained on datasets from Abu Dhabi and San Francisco. Results show a 20% reduction in mean absolute error compared to existing methods, demonstrating potential for scalable coastal flood management.
MASARAT SA has developed Mubeen, a proprietary Arabic language model specializing in Arabic linguistics, Islamic studies, and cultural heritage. Mubeen was trained using native Arabic sources, including digitized historical manuscripts processed via a proprietary Arabic OCR engine. The model employs a Practical Closure Architecture to improve user intent understanding and provide decisive guidance. Why it matters: Mubeen addresses the utility gap in current Arabic LLMs by focusing on native Arabic data and cultural authenticity, which is critical for heritage preservation and alignment with Saudi Vision 2030.
Professor Peter Goadsby, a neurologist and neuroscientist, has been appointed as Senior Associate to the President and Founding Dean of KAUST's new Division of Biomedical Sciences. He will lead the establishment of the university's fourth academic division, focusing on Biomedical Sciences, and advance the neuroscience department. Goadsby's research identified CGRP as a central driver of migraine, leading to new medicines and earning him the 2021 Brain Prize. Why it matters: This appointment strengthens KAUST's and Saudi Arabia's capacity to translate research into healthcare solutions and supports the Kingdom’s Vision 2030 goals in health innovation.
A KAUST-led study has revealed a unique microbial ecosystem in the Hatiba Mons hydrothermal vent fields of the Red Sea, first documented in 2023. Using genome-resolved metagenomics, the study reconstructed over 300 microbial genomes from five vent sites. The analysis showed an ecosystem dominated by microbes capable of iron, sulfur, nitrogen, and carbon cycling, unlike most hydrothermal vents that are sulfur- and methane-based. Why it matters: The discovery provides new insights into microbial processes in extreme conditions, ocean resilience, and global carbon cycling, highlighting the interplay between geology and biology in the Red Sea.
KAUST professors Roberto Di Pietro and Marc Dacier co-authored a paper on cybersecurity strategies for Saudi Arabia and the Arab world, published in Communications of the ACM. The paper outlines a multidisciplinary framework for digitization aligned with Saudi Vision 2030, emphasizing global best practices, cultural adaptation, and capacity building. KAUST is positioned to advise on national cybersecurity policy in cooperation with the Saudi National Cybersecurity Authority. Why it matters: The framework addresses the critical need for advanced cybersecurity to support Saudi Arabia's rapidly growing digital economy and infrastructure.
This survey paper analyzes over 40 benchmarks used to evaluate Arabic large language models, categorizing them into Knowledge, NLP Tasks, Culture and Dialects, and Target-Specific evaluations. It identifies progress in benchmark diversity but also highlights gaps like limited temporal evaluation and cultural misalignment. The paper also examines methods for creating benchmarks, including native collection, translation, and synthetic generation. Why it matters: The survey provides a comprehensive reference for Arabic NLP research and offers recommendations for future benchmark development to better align with cultural contexts.
This paper presents the development and validation of an Arabic version of the Attitudes Toward Large Language Models (AT-GLLM and AT-PLLM) scales, adapted from the original English versions. The study involved translating the scales and testing them on a sample of 249 Arabic-speaking adults. The translated scales demonstrated strong psychometric properties, including a two-factor structure, measurement invariance across genders, and good reliability and validity. Why it matters: This provides a culturally relevant tool for assessing attitudes toward LLMs in the Arab world, crucial for localized research and policy-making in the rapidly growing field of Arabic AI.
KAUST and Aramco have signed an IP license agreement for the Hybrid Unmanned Aerial Vehicle Crawler Technology (HUCT), which combines aerial and terrestrial robotic capabilities. The HUCT will be used for inspection, monitoring, and other industrial applications within Aramco's operations. The agreement supports KAUST's mission to translate research into real-world applications, aligned with Saudi Vision 2030. Why it matters: This partnership demonstrates the growing innovation ecosystem in Saudi Arabia, leveraging academic research for industrial applications in robotics and AI.
KAUST researchers have found conclusive evidence that the Red Sea completely dried out approximately 6.2 million years ago. Using seismic imaging, microfossil evidence, and geochemical dating, they determined a massive flood from the Indian Ocean refilled it in about 100,000 years. The flood carved a 320-kilometer-long submarine canyon and restored marine conditions. Why it matters: This discovery provides insights into extreme environmental events and the Red Sea's unique geological history, distinguishing it from the refilling of the Mediterranean.
Researchers introduce MATRIX, a vision-centric agent tuning framework for robust tool-use reasoning in VLMs. The framework includes M-TRACE, a dataset of 28.5K multimodal tasks with 177K verified trajectories, and Pref-X, a set of 11K automatically generated preference pairs. Experiments show MATRIX consistently outperforms open- and closed-source VLMs across three benchmarks.
Deloitte Middle East and KAUST have signed an MoU to collaborate on AI applications in Saudi Arabia. The partnership aims to bridge the gap between AI research and business applications, aligning with Saudi Vision 2030. Joint projects will address AI challenges, develop talent through training programs, and focus on knowledge transfer and ethical AI governance. Why it matters: This collaboration accelerates AI innovation in Saudi Arabia by combining global expertise with local knowledge, fostering economic and societal impact.
KAUST researchers found that wildfire smoke particles act as chemical factories under sunlight, producing harmful oxidants like peroxides. These particles bypass traditional suppression by nitrogen oxides in polluted environments, generating oxidants internally. The study reveals that colored organic molecules in biomass-burning aerosols act as photosensitizers, triggering rapid reactions. Why it matters: The findings highlight that current air-quality and climate models underestimate oxidant production from wildfires, with implications for anticipating health risks and environmental impacts in regions like Saudi Arabia.
Kuwait has successfully conducted the farthest remote robotic surgery, an achievement recognized by Guinness World Records. This pioneering medical procedure showcased advanced capabilities in tele-operation and robotic technology within the healthcare sector. The specific details of the technology used and the medical context were highlighted in the record-setting event. Why it matters: This accomplishment signifies a major leap in medical innovation and remote healthcare capabilities for Kuwait and the broader Middle East region.
KAUST researchers are using CarboSoil biochar and native biocrusts to revitalize arid lands in Saudi Arabia, enhancing soil fertility, capturing carbon, and reducing erosion. CarboSoil, engineered from poultry waste by KAUST's Himanshu Mishra, improves nutrient and water retention in desert soils. Terraxy, Mishra's startup, aims to convert all of Saudi Arabia's poultry waste into CarboSoil, supporting greening initiatives. Why it matters: This technology offers a sustainable solution to boost domestic food production, combat desertification, and reduce landfill waste in Saudi Arabia, aligning with the Kingdom's food security and environmental goals.
The paper introduces TimeHUT, a new method for learning time-series representations using hierarchical uniformity-tolerance balancing of contrastive representations. TimeHUT employs a hierarchical setup to learn both instance-wise and temporal information, along with a temperature scheduler to balance uniformity and tolerance. The method was evaluated on UCR, UAE, Yahoo, and KPI datasets, demonstrating superior performance in classification tasks and competitive results in anomaly detection.
A Heritage Commission and KAUST collaboration published in Nature Communications reveals the discovery of large-scale rock art panels in the Nefud Desert, dating back 12,000 years. Over 60 panels with 176 engravings were found depicting animals like camels and ibex. Paleoenvironmental analysis indicates surface water was present 14,000 years ago, supporting early human and wildlife expansion. Why it matters: The findings revise the timeline of human repopulation in Saudi Arabia's interior deserts after the Last Glacial Maximum and demonstrate the significance of interdisciplinary research in understanding the region's climate history.
Aramco, Honeywell, and KAUST have signed a Joint Development Agreement (JDA) to co-develop a next-generation direct Crude-to-Chemicals (CTC) technology. The collaboration aims to develop and scale up the full CTC process, reducing capital and operating costs. The new CTC pathway is designed to convert crude oil directly into light olefins and other high-demand chemicals. Why it matters: This partnership signifies a major push for Saudi Arabia to lead in downstream innovation, aligning with Vision 2030 to diversify the economy and strengthen its position in the petrochemicals industry.
Researchers introduce ALARB, a new benchmark for evaluating reasoning in Arabic LLMs using 13K Saudi commercial court cases. The benchmark includes tasks like verdict prediction, reasoning chain completion, and identification of relevant regulations. Instruction-tuning a 12B parameter model on ALARB achieves performance comparable to GPT-4o in verdict prediction and generation.
The study analyzes over 1,000 images generated by ImageFX, DALL-E V3, and Grok for 56 Saudi professions, finding significant gender imbalances and cultural inaccuracies. DALL-E V3 exhibited the strongest gender stereotyping, with 96% male depictions, particularly in leadership and technical roles. The research underscores the need for diverse training data and culturally sensitive evaluation to ensure equitable AI outputs that accurately reflect Saudi Arabia's labor market and culture.
A KAUST-led study tracked clownfish and anemones in the Red Sea from 2022-2024, finding that extreme heat caused anemone bleaching, followed by near-total clownfish death, and then anemone death. The heatwave saw accumulated thermal stress reach 22 degrees heating weeks, far exceeding the threshold for coral bleaching. The research highlights heat risks faced by non-coral reef organisms and the need for taxon-specific thresholds to predict risks to reef symbiotic relationships. Why it matters: The Red Sea is a bellwether for climate change impacts on marine ecosystems, and this study underscores the urgency of conservation efforts like KAUST's Coral Restoration Initiative.
The Hala technical report introduces a family of Arabic-centric instruction and translation models developed using a translate-and-tune pipeline. A strong Arabic-English teacher model is compressed to FP8 and used to create bilingual supervision data. The LFM2-1.2B model is fine-tuned on this data and used to translate English instruction sets into Arabic, creating a million-scale corpus. Why it matters: The release of models, data, evaluation tools, and recipes will accelerate research and development in Arabic NLP, providing valuable resources for the community.
KAUST researchers are developing passive cooling solutions that use no electricity to address Saudi Arabia's high air conditioning electricity consumption. The technologies leverage nanotechnology, reflective materials, water evaporation, and advanced sensors to cool urban spaces, greenhouses, and buildings. One innovation involves nanotechnology that absorbs water from the air to cool electronics. Why it matters: These advancements are crucial for sustainable growth in hot climates, particularly for protecting solar panel efficiency and addressing rising global energy demands for cooling.
Researchers address the challenge of limited Arabic medical dialogue data by generating 80,000 synthetic question-answer pairs using ChatGPT-4o and Gemini 2.5 Pro, expanding an initial dataset of 20,000 records. They fine-tuned five LLMs, including Mistral-7B and AraGPT2, and evaluated performance using BERTScore and expert review. Results showed that training with ChatGPT-4o-generated data led to higher F1-scores and fewer hallucinations across models. Why it matters: This demonstrates the potential of synthetic data augmentation to improve domain-specific Arabic language models, particularly for low-resource medical NLP applications.
KAUST researchers have developed a system to convert captured carbon dioxide into industrial-grade ethylene using a high-pressure electrolyzer. The system operates under realistic industrial conditions and uses captured, high-pressure CO₂. It reduces the energy cost of producing ethylene by 0.8 gigajoules per metric ton compared to existing electrolysis systems. Why it matters: This innovation presents a direct path for transforming greenhouse gas emissions into valuable chemical products, aligning with Saudi Arabia's circular economy goals.