Research applying artificial intelligence across science and engineering. Covers machine learning, sensing and quality control, immune and health prediction, and 5G networks.
Researchers developed a new analysis method, LM-QASAS, combining antibody language models with B-cell receptor repertoires to identify candidate sequences responding to an immune stimulus. The method successfully tracked the rise, peak, and decline of immune responses in new individuals, even for emerging infectious diseases.
A new study by KTU researchers explores whether artificial intelligence can uncover patterns in glucose data to predict changes and assess hypoglycaemia risk. The model achieved high accuracy in forecasting glucose levels and performing well on patients with unseen data.
Researchers developed a new mathematical framework to test AI transparency in clinical settings, detecting information leakage in concept-based AI models. The framework provides practical guidance for designing models that minimize leakage and offer meaningful transparency.
The University of North Florida has received a grant from the Sloan Foundation to develop generative AI tools for faster and more reliable research software code updates. The goal of the project is to help evolve software used by researchers globally while preserving accuracy and reliability.
The joint factsheet highlights the complementary activities of FORSAID, CERBERUS and STELLA, demonstrating how innovative digital technologies can improve early detection and management of plant pests. The projects share a vision of using AI, remote sensing and citizen engagement to transform Europe's response to pest outbreaks.
The article highlights the need for governance addressing the needs of historically marginalized populations to mitigate AI deployment risks. The authors recommend equity impact assessments, transparency, and sustained investment for AI in public health to protect underserved communities.
Researchers at Memorial Sloan Kettering Cancer Center used AI to design better CAR T cell therapies by recognizing specific proteins on cancer cells. The AI approach outperformed existing binders in a proof-of-concept study, leading to a better understanding of why some cancer-targeting proteins work well and others fail.
A new partnership will provide access to a constantly evolving training environment, enabling the development of practical cyber skills and strengthening connections between education, research, industry, and defence. The partnership aims to build Australia's sovereign cyber capability and ensure resilience in the face of evolving cybe...
As large crowds gather for the funeral, authorities prepare for potential cyberattacks, including phishing, disinformation campaigns, and denial-of-service attacks. Experts like Basel Katt believe the risk of attacks increases due to the event's attention and resources.
Researchers developed Neural Value Alignment (NVA), a brain-computer interface technology that detects cognitive mismatches between humans and AI through brainwaves. The AI system can revise its actions in real time according to human goals, accelerating the shift from explicit command-based AI to inferential intent-based AI.
The authors propose norms for AI use in research, distinguishing between non-generative tasks and generative AI, which should not replace human interpretation and judgement. Science requires diversity of inputs and ideas, and serendipity counts, making generative AI training on existing data limiting breakthroughs.
Researchers developed a physics-based AI approach to predict global-scale carbon cycling in ocean sediments, resolving a long-standing challenge in climate science. The study reveals key findings on dissolved organic carbon behavior, including 11% of particulate organic carbon returning to seawater.
The SMASH Team, led by Professor Ping LUO, won the silver medal in robot table tennis at the World Humanoid Robot Games. The team's autonomous humanoid table tennis system, SMASH, is the world's first to use onboard sensing to achieve consecutive strikes and rallies with human players.
Professor Wang joins HKU as Chair Professor, bringing his pioneering research on the prefrontal cortex and computational psychiatry. His work will foster interdisciplinary collaborations and advance computational models for psychiatric care.
The HKU Robocon team, consisting of 40 students, won the Asia-Pacific Broadcasting Union's Asia-Pacific Robot Contest 2026, defeating teams from 15 countries. The team's exceptional collaboration, innovative engineering, and operator training contributed to their victory.
USC researchers will lead a $20M NSF effort to develop AI-powered optimization tools for power grids and supply chains. The project aims to improve decision-making in these complex systems and expand AI education for high school students.
George Mason University has selected three companies for its 2026 Global Scale-Up Campus 'Incheon Univ. X' initiative. Massimiliano Albanese will provide advisory services for the program, supporting U.S. activities. Funding for Albanese's services was provided by Mason Korea IUCF and will last from August 2026 to November 2026.
Researchers Ahmedullah Aziz and Sai Swaminathan received NSF CAREER Awards for their projects in superconducting logic systems and AI for community impact. Aziz will develop superconducting electronics for high-performance computing, while Swaminathan is creating low-cost AI devices for local problem-solving.
A team from Helmholtz Munich developed MemBrain v2, an AI tool that automates the analysis of cell membranes in 3D images, cutting work from weeks to hours. The tool locates specific membrane proteins and analyzes their spatial arrangement, showing how cellular processes are organized at the molecular level.
The ACC Latin America 2026 conference will bring together experts and cardiovascular clinicians to examine emerging cardiovascular science and evidence-based strategies for improving global heart health. The conference will also recognize individuals who have earned the distinguished Fellow of the ACC (FACC) designation.
A new AI tool has been developed to rapidly and accurately measure colonoscopy quality, improving care quality and reducing colorectal cancer mortality. The tool analyzed nearly 19,000 colonoscopies performed by 55 physicians, identifying key moments and tracking quality indicators that cannot be feasibly measured by humans at scale.
An international team proposes a physics-aware framework to make AI-guided hydrogen storage materials discovery more reliable. The framework connects reproducibility-aware data, thermodynamics-constrained models, AI-driven inverse design, and experimental validation to create a learning cycle.
Researchers have developed the AlphaGenome Atlas, a comprehensive map of more than 9 billion possible single-letter DNA changes. The one-petabyte dataset provides artificial intelligence-generated predictions for the molecular effects of these changes, accelerating understanding of the human genome.
A new study reveals that AI engineers are aware of ethical risks but face structural barriers that prevent them from acting on them. The researchers found that engineers want to implement safeguards but feel unable to do so due to organisational factors such as tick-box compliance processes and commercial pressures.
Generative AI can help travellers organise their thoughts, compare options and complete holiday plans, making it easier for those who find planning overwhelming. The technology can provide structure and support, freeing travellers to focus on making judgements and decisions.
A recent study found that AI chatbots incorrectly reassure sleep apnea patients that their symptoms are not serious, discouraging them from seeking specialist assessment. This can lead to delayed diagnosis and treatment, increasing the risk of complications such as high blood pressure, stroke, and heart disease.
Recent health tech advancements include drone delivery for organ transportation, AI tools in education with potential risks to childhood development, and early cancer neurotechnology research. Consumer wearables are shifting towards data interpretation, with wearable platforms needed to provide meaningful insights from collected data.
Researchers developed an AI tool, AIMe, to predict and organize small molecule structures, covering over 100 million compounds. The tool, called DeepMS2Reasoner, uses neuro-symbolic AI to simulate molecule fragmentation, providing interpretable outputs and accelerating hypothesis generation.
The UN University's latest publication highlights the need for domain-informed AI in grid planning to address physical and fiscal risks from outdated climate data. The authors warn that 15-20 year lifespans of electricity infrastructure are based on historical weather records unlikely to hold in the coming decades.
Researchers at the University of Arizona assessed seven generative AI models for fallibility, persuadability and correctability during lengthy conversations. The study revealed intrinsic limitations of these models, including oscillation between accepting and rejecting false statements, and the need for careful human engagement to miti...
Dr. Sophie de Vries receives funding to study how plants balance immunity with cooperation, while Dr. Tristan Stöber works on developing AI systems that can build accurate internal models of the world. Professor Elisa Oberbeckmann investigates gene regulation mechanisms.
A new report from the American Psychological Association warns that student engagement with EdTech is not the same as learning, and that generative AI tools pose particular risks. The report recommends prioritizing tools that build durable knowledge and skills, and exercising caution when introducing AI tools into learning environments.
A new probabilistic method, Bayesian Probabilistic Data Association via Gaussian Mixture Models, improves trajectory accuracy and semantic mapping quality in robots. The approach reduces duplicate registrations and handles ambiguous observations, enabling more stable and reliable object-level maps.
The AI BioDesign accelerator will generate open models, datasets, and tools to create new biological solutions for human health and environmental challenges. The goal is to learn and model the rules biology uses to build life, enabling the development of new drugs, enzymes, and biological computers.
ISTA researchers have secured 5 ERC Starting Grants to fund innovative projects in AI, astrophysics, and optics. The grants will support early-career researchers in understanding bacterial immune systems and developing new microscope techniques to study elusive quantum matter.
Researchers have mapped the complete neural wiring behind taste in an animal, tracing taste signals from sensory neurons to motor neurons that drive feeding. The study reveals neural pathways that could explain how taste triggers anticipatory insulin release before a single calorie is absorbed.
The complete connectome of the fruit fly brain has been mapped, revealing new insights into how the brain enables sophisticated behavior. The 166,000-neuron map, developed through Janelia Research Campus's pioneering efforts, has transformed scientific research and pioneered new technologies.
Computational linguist Michael Hahn aims to improve AI reasoning in large language models, particularly in handling interdependent sequences and distinguishing between similar pieces of information. He plans to develop a theory to explain how training conditions affect AI's ability to develop reliable conclusions.
The new AI data center at KIT will provide high-performance computing power to support research in medicine, sustainable energy, and modern AI applications. The data center will use waste heat to heat buildings on campus, optimizing operations and strengthening sustainability efforts.
Parth Pathak from George Mason University aims to create an AI-assisted cross-layer timing anomaly detection framework to secure data center timing infrastructure. The project received $100,000 in funding from the Virginia Innovation Partnership Authority.
The 12 companies selected will work alongside Mayo Clinic Platform's experts and leverage its de-identified clinical data ecosystem to advance digital health solution development. The program aims to transform healthcare through AI-enabled solutions leveraging clinical data and advanced analytics.
A Stanford-led team created a quantum-optical spin glass to increase AI memory capacity. The network, called a quantum-optical spin glass, has a greater capacity to hold and recall memories than traditional AI networks, exhibiting short-term plasticity similar to the brain's synaptic connections.
A review led by UCLA investigators found that AI can identify subtle signs of breast cancers missed during routine mammograms, but its clinical benefit is uncertain. AI may also identify patterns associated with an increased risk of cancer before it becomes visible on a mammogram.
Organic chemistry is challenging AI researchers to think outside the box, driving advances in how AI represents complex problems and reasons from limited evidence. This is enabling the development of more reliable, efficient, and transparent AI systems that can support various areas of science and society.
AI has evolved from an ambitious idea to a powerful force in science, medicine, industry, and environmental protection. Emerging co-scientist systems can generate hypotheses, evaluate ideas, and refine research directions, pointing toward a future of AI as a digital scientific collaborator.
Jack Dongarra, Yilu Liu, and Parans Paranthaman receive R&D 100 Awards for their work on innovations in large-scale computing and power grid resilience. Dongarra's team developed Fenix, an open-source software that detects and repairs hardware failures in supercomputing applications.
KAIST researchers develop SafeQL, a technology that identifies and selectively corrects errors in AI-generated SQL queries, reducing the need to regenerate entire queries. The technology improves data retrieval accuracy and speed, accelerating the adoption of AI work assistants in enterprise environments.
A scoping review found that only 77 studies validating 52 medical AI products reported demographic data, highlighting the need for more transparent reporting to confirm safe and unbiased performance. The review also showed that performance reporting for demographic subgroups is inadequate, posing a risk to patient care and outcomes.
A new framework jointly optimizes UAV trajectories and FANET topology to maximize data transmission, outperforming existing methods in field experiments and simulations. The approach enables more efficient and reliable multi-UAV missions for environmental monitoring and other applications.
The PACMAN AI framework successfully tested in five real-world experiments, making decisions in milliseconds and surpassing human reaction time. The framework combines multiple machine learning models to monitor and control different aspects of the fusion system, enabling real-time control and safety.
A new AI framework, Perspective, provides a structured approach to explaining complex patterns in AI predictions, enabling researchers to test hypotheses and improve designs. By revealing the underlying relationships, XAI can support discovery, optimisation, and certification for AI in high-stakes fields.
The congress featured over 1,750 sessions and 12 Hot Line sessions highlighting practice-changing findings in cardiovascular science. ESC Congress 2026 also released three clinical practice guidelines and the Fifth Universal Definition of Myocardial Infarction.
A study found that older adults prefer chatbots that are authentic and responsive, rather than human-like, to provide companionship and support. The chatbot's ability to remember past conversations and adapt to user preferences is crucial for its success.
Researchers propose an AI framework, called interoceptive AI, that uses internal states to inform learning and decision-making in dynamic environments. This approach treats internal conditions as a continuous source of context, influencing what an agent learns, prioritizes, and does.
A new method, called CW-Net, translates the reasoning process of an autonomous vehicle's AI system into understandable concepts that explain its behavior. CW-Net explains the decisions of machine learning-based planners using concepts like
A new AI-assisted approach uses artificial intelligence to help novice users capture heart images and identify patients with aortic stenosis. The approach, presented in a study published in JAMA Cardiology, enables novice users to collect focused heart ultrasound images and analyze them for signs of moderate or greater aortic stenosis,...
Researchers studied kids playing with AI toys, finding they exhibited curiosity, but also frustration and antipathy when toys failed to respond correctly. The toys' inability to engage in complex conversations led to kids testing their limits and even antagonizing them.
KAIST researchers have developed a new catalyst that can remove tetrafluoromethane (CF₄), a greenhouse gas 6,000 times more potent than CO₂, with high efficiency. The catalyst, called entropy-stabilized aluminate (ESA), harnesses the power of disorder to stabilize its structure and maintain performance over extended periods.
The Digital Science Catalyst Grant 2026 seeks novel applications of agentic AI workflows that can benefit any part of the research lifecycle, including areas such as data management and knowledge discovery. The grant aims to support early-stage solutions building research agents with provenance, governance, and accountability.
The authors propose a spectrum of clinical autonomy, ranging from advisory tools to navigator systems that work with minimal human oversight, to enable early prediction and prevention of disease. However, challenges related to clinical validation, integration, data quality, and regulatory approval limit widespread deployment.