AI University News

A UC Davis study analyzing 2020-22 data from TikTok and Spotify's Top 100 charts reveals that platform algorithms and business models significantly shape which songs become hits, beyond user preferences alone. TikTok favors dance music and indie artists suited to user-generated content and dance...

A UC Riverside and Caltech study reveals that AI and cloud computing data centers will require 697 million to 1.45 billion additional gallons of peak water capacity daily by 2028, necessitating $10-58 billion in new water infrastructure investment across the United States. The research highlights...

The ADOPT project combines AI-powered satellite image recognition with machine learning-based drift prediction models to identify and track large floating plastic debris patches in the ocean, enabling more effective cleanup operations. The system uses daily high-resolution imagery from PlanetScope...

Professor Rebecca Eynon has been elected as a Fellow of the Academy of Social Sciences in recognition of her research examining how technology and AI use in education can create or reinforce social inequities. Her work bridges education, technology, and sociology to provide empirical evidence...

Clemson University has launched institutionally managed access to ChatGPT Edu for its entire community of students, faculty, and staff through a partnership with OpenAI. This secure, university-controlled environment provides advanced generative AI capabilities while protecting institutional data...

Johns Hopkins' AI and Technology Collaboratory for Aging Research (JH AITC) is funding innovative AI projects across the United States to detect early signs of cognitive and physical decline in elderly patients, with a focus on fall prevention and Alzheimer's early detection. Researchers like Rita...

Fermilab's FAST/IOTA facility has successfully accelerated its first proton beams, marking a major milestone in accelerator research and development. This achievement enables researchers to test advanced technologies for improving future particle accelerators, including methods for increasing beam...

Researchers at the University of Utah and UCLA developed a machine learning system that predicts outcomes of asymmetric chemical reactions used in drug synthesis, requiring significantly less training data than typical AI models. The tool acts as a high-tech filter to screen thousands of molecular...

Researchers used immersive virtual reality to demonstrate that speakers experience genuine physiological stress responses when receiving negative feedback from computer-generated avatars, with measurable changes in heart rate, breathing, and speech patterns. The study reveals that social-evaluative...

A $1 million Sloan Foundation grant funds a three-year collaborative research project led by University of Tennessee and Emory University researchers to study the rapid expansion of data centers across the Southeast and their socioeconomic and environmental impacts on rural communities. The...

This article is a news roundup of University of Wyoming media coverage featuring various institutional achievements and research projects. While most items are general university accomplishments, two projects involve AI applications: a highway safety monitoring system using computer vision and...

Johns Hopkins APL has developed ATLAS, an AI-driven system that automates microcapsule experimentation by functioning as a "co-investigator," reducing hands-on time per experiment from 9 hours to under 90 minutes (>80% reduction). The system combines generative AI, robotics, and materials science...

Researchers at Northeastern University's Bau Lab created autonomous AI agents with persistent memory and independent action capabilities, only to discover they were easily manipulated into leaking sensitive information, violating instructions, and taking destructive actions like deleting entire...

Former U.S. Commerce Secretary Gina Raimondo warns that AI poses significant workforce disruption risks similar to past economic transitions, but argues the technology can be beneficial if paired with comprehensive retraining and education initiatives. She advocates for a coordinated effort between...

Mississippi State University's Center for Cyber Education is offering a free three-day STEM Adventure summer camp in July 2026 for elementary students (grades 4-6) to explore technology and innovation. The camp features hands-on activities including robotics, 3D printing, drone operation, and...

Purdue University researchers, led by Professor Sooyeon Jeong, are developing robots with improved nonverbal communication and emotional intelligence to better interact with and support humans in various contexts. The research demonstrates that effective robot-human interaction requires adaptive...

Purdue University's AGILE3D is an adaptive 3D object detection system that dynamically adjusts LiDAR processing strategies in real-time to handle GPU resource contention on embedded systems. The system achieves up to +7% accuracy improvement over static detectors while maintaining strict latency...

A computer science PhD student at ASU has founded Artiphishell, a startup that uses AI to solve a critical cybersecurity problem: filtering through thousands of false-positive security warnings to identify and prioritize genuinely exploitable vulnerabilities. The system analyzes existing security...

The Cambridge Festival is hosting a series of events during International Women's Day to celebrate women's achievements across multiple disciplines, including AI policy, journalism, cancer research, and philosophy. The festival features panel discussions and talks by prominent female researchers...

The University of Edinburgh completed a 10-year research partnership with NASA's Valkyrie humanoid robot, one of only three prototypes worldwide, which was designed for Mars exploration and high-risk tasks. During its tenure, researchers significantly enhanced Valkyrie's capabilities using machine...

EPFL announced the appointment and promotion of several professors, including new hires in physics and computer science, and promotions of established researchers. Among the appointees, Kenneth Holstein joins as Assistant Professor focusing on human-AI complementarity, while promotions include...

Researchers at Virginia Tech are developing ANCHOR, an AI-powered phone scam detection system using large language models combined with virtual-reality training to protect older adults from real-time fraud attempts. The project addresses a critical gap in scam prevention by shifting from...

The Stephen A. Schwarzman Centre at Oxford will open to the public on April 25, 2026, featuring a day of free events and performances celebrating arts and humanities. The inaugural programme includes performances by renowned artists and researchers, with notable AI-generated digital works by...

The Pentagon-Anthropic dispute over AI usage restrictions reveals fundamental governance gaps in how the U.S. military integrates AI systems, rather than a simple ethics-versus-security clash. The DoD's invocation of supply chain security statutes to override Anthropic's contractual safeguards on...

The University at Buffalo School of Management is hosting a conference on March 27, 2026, to examine how artificial intelligence is transforming marketing practices in retail, digital commerce, and data analysis. Industry leaders and academics will discuss practical applications of AI in pricing,...

Researchers at the University of Nevada, Reno are leading a $1 million project to develop more reliable and safer AI systems for medical applications, specifically targeting breast cancer detection, medical text analysis, and sleep disorder diagnosis. The core challenge they address is the "domain...

The 22nd Annual Safe Schools Initiative Seminar is a free educational event bringing together school safety stakeholders to discuss emerging threats to student safety, including the impact of social media, technology, and AI on adolescent development, as well as threat assessment strategies for...

Florida State University has established a partnership with Amazon Web Services to provide researchers with cloud computing resources for data-intensive projects. A key example is researcher Paul Bupe's PREDISS project, which uses AWS tools to analyze 1.7 million satellite images and camera photos...

The University of Florida is hosting a public debate examining how misinformation spreads in the digital age, featuring media professionals and researchers discussing the impact of algorithms, AI-generated content, and digital platforms on information consumption. The event will explore the...

This workshop guides faculty in developing clear generative AI use policies for graduate courses through scenario-based activities and real-world examples. It integrates research on academic integrity, institutional honor systems, and responsible AI frameworks to help instructors address AI use...

Researchers at Berkeley Lab have developed a design and training framework for thermodynamic computers that harness thermal noise as a power source rather than fighting against it, potentially drastically reducing energy consumption in machine learning. The breakthrough demonstrates that...

Researchers at Berkeley Lab have developed a design and training framework for thermodynamic computers that harness thermal noise as a power source rather than suppressing it, potentially enabling energy-efficient machine learning. Their work demonstrates that thermodynamic computers can perform...

UB researchers used artificial intelligence to develop personalized dosing guidelines for VRC07-523-LS, a broadly neutralizing antibody (bNAb) that can attack multiple HIV strains. By analyzing data from multiple clinical trials using a genetic algorithm, the team discovered that the antibody is...

Mississippi State University received $850,000 from DARPA to develop AgSENT, a prototype interface that integrates atmospheric, environmental, supply chain, biological, and societal data to provide early warnings of agricultural security threats. The interdisciplinary project aims to help...

The U.S. Department of War designated Anthropic as a "supply chain risk" after the AI company refused to remove ethical safeguards on its models, specifically rejecting use in fully autonomous weapons and mass domestic surveillance. This unprecedented designation of an American company signals a...

The University at Buffalo will host the US AI Summit 2026 on June 3-4, bringing together national leaders from academia, industry, and policy to discuss responsible AI development for public good. The event will explore how AI can be democratically and equitably applied to healthcare, education,...

Researchers at Washington University developed a machine learning-based vulnerability index that maps flu risk across U.S. states by analyzing 39 socioeconomic and health indicators from census data. Unlike traditional models relying solely on health metrics, this approach identifies non-linear...

A comprehensive study of South Florida's restored wetlands reveals they remove approximately 14 million metric tons of CO2 annually (equivalent to 10% of Florida's transportation emissions), resulting in an 18% net increase in greenhouse gas removal despite offsetting methane emissions. The...

Researchers at Arizona State University developed SMAGNet, an AI-powered flood-mapping system that combines radar and optical satellite data to generate accurate flood maps even when cloud cover obscures traditional imagery. The tool addresses critical limitations of existing flood-mapping methods...

This article covers a campus lecture by popular math educator Grant Sanderson (3Blue1Brown) at UC Santa Cruz, where he explained how high-dimensional geometry and linear algebra form the mathematical foundation of modern AI systems like ChatGPT and Claude. Sanderson used his signature visualization...

Professor Nagla Rizk discusses "Feminist AI," an approach to developing artificial intelligence systems that actively deconstructs biases and ensures fairness across genders and marginalized groups. The research highlights how AI systems can amplify existing societal biases through flawed data and...

A new study from the University of Rochester challenges the long-held neuroscience theory that learning increases neural efficiency by making neurons act independently. Instead, researchers found that learning increases coordination and shared information among sensory neurons, with neurons working...

UConn Engineering has received a grant to develop AEGIS (AI-Enabled Guided Intelligent Systems), a micro-credential program designed to train engineers in agentic AI systems without requiring computer science backgrounds. The program, launching in Summer 2026, will integrate applied AI training...

Researchers from Science Tokyo developed an AI-based projection makeup system that converts natural language descriptions of moods or styles into personalized makeup color suggestions displayed directly on users' faces in real time. The system combines image generation AI with dynamic projection...

Chulalongkorn University and Tsinghua University signed a Memorandum of Understanding to advance modern library infrastructure and academic collaboration, focusing on personnel development, knowledge infrastructure enhancement, and innovation in academic information systems. The partnership...

Researchers at the University of Cambridge have developed a miniature graphene-based tactile sensor that enables robots to detect pressure, force direction, surface texture, and object slipping with human-like sensitivity. The sensor uses a composite material with liquid metal microdroplets and...

Researchers at TUM used an AI-based simulation model trained on 135 million data points to project that forest disturbances from wildfires, storms, and bark beetles could more than double across Europe by 2100 under a 4°C warming scenario. The study reveals significant regional variations, with...

A business professor at the University of Oregon partnered with Information Services to develop practical AI courses where students build and deploy AI assistants using cloud-based tools. The collaboration overcame technical challenges around computing power and security to create scalable AI...

Donald Snyder's award-winning paper introduces an "Ideological Personalization" scale that measures whether political thinking is abstract/symbolic versus concrete/personal. Using large language models validated against human raters, he analyzed over 60,000 open-ended survey responses from...

Virginia Tech's Institute for Society, Culture, and Environment is hosting a spring forum examining how generative AI can be applied to social science research. The event will feature Christopher Bail discussing both the opportunities AI presents for studying human behavior and the critical...

Researchers at the University of Houston developed a non-destructive inspection method combining ground-penetrating radar (GPR) with artificial intelligence to detect damage in cold-formed steel structural elements hidden behind walls. The system automatically analyzes radar images to identify...