AI News

RSS

Five Minutes of Training Helps People Detect AI-Generated Fake Faces

A study by researchers from the Universities of Reading, Greenwich, Leeds, and Lincoln found that just five minutes of targeted training can markedly improve people’s ability to spot AI-generated faces. After training, accuracy rose from below chance levels to 64% for super-recognisers and 51% for typical participants.

11 Nov 2025

Chiba University Researchers Develop Topology-Aware AI To Decode Brain Signals For Thought-Controlled Devices

Researchers at Chiba University have developed the Topology-Aware Multiscale Feature Fusion (TA-MFF) network, an advanced deep learning model that enhances the decoding of motor imagery EEG signals for brain-computer interface (BCI) applications. By integrating spectral, topological, and spatiotemporal features, the system achieves superior accuracy in interpreting neural activity linked to imagined movement.

11 Nov 2025

Tufts Medical School Leads the Way in Training Future Doctors to Use AI Responsibly

A Tufts University School of Medicine study led by Maria Blanco found that while two-thirds of faculty and students see potential in AI, fewer than 12% feel proficient. The findings are shaping Tufts’ roadmap for responsible AI integration into medical education through training, policy updates, and hands-on pilot projects.

11 Nov 2025

Biased AI Steers Human Hiring Decisions, UW Study Finds

A University of Washington study with 528 participants shows that people mirror biased large language model recommendations during hiring, even when applicants are equally qualified. Without source bias, selections were balanced, but with moderately to severely biased AI, human choices aligned closely with the AI, highlighting risks for fairness in recruitment.

11 Nov 2025

Penn GSE Launches First High School Curriculum to Teach Students How to Audit AI Bias

The University of Pennsylvania’s Graduate School of Education (Penn GSE) has launched AI Auditing for High School, a first-of-its-kind curriculum that empowers students to identify and analyze algorithmic bias in AI systems through real-world audits, no coding required. Supported by national foundations and educators, the program aims to foster critical digital literacy and ethical AI awareness.

11 Nov 2025

South Korean Scientists Develop AI to Detect Hidden Cracks in Structures With Unprecedented Precision

Researchers at Chung-Ang University in South Korea have developed DiffectNet, an AI-powered diffusion model that reconstructs hidden internal defects in structures with high precision, overcoming the physical limits of traditional non-destructive testing. The breakthrough enables real-time defect imaging for critical sectors such as energy, aerospace, and semiconductors.

11 Nov 2025

NTU and Zero Gravity Launch SG$5 Million Hub to Build Decentralised AI Systems

Nanyang Technological University, Singapore (NTU) and Zero Gravity (0G) have launched a S$5 million joint research hub to advance blockchain-based AI technologies that promote transparency, accountability, and decentralisation. The four-year initiative will pioneer open AI systems, secure marketplaces, and proof-of-useful-work frameworks.

10 Nov 2025

Scientists Use Explainable AI to Reveal How Sensors Recognize Smells

Researchers at Japan’s National Institute for Materials Science (NIMS) have used explainable AI (XAI) to uncover how chemical sensors distinguish between odorant molecules, revealing the molecular interactions that drive artificial smell recognition. The findings pave the way for smarter sensor design and deeper insights into human olfaction.

10 Nov 2025

Tokyo Scientists Use AI to Decode Material Structures From X-Ray Spectra

Researchers at Tokyo University of Science have developed an AI-based method to automatically analyze X-ray absorption spectroscopy (XAS) data, achieving precise and objective identification of material structures and defects. Using machine learning, particularly the UMAP algorithm, the approach enables rapid, scalable, and noise-resistant characterization of boron-based materials.

10 Nov 2025

Study Finds Hidden Bias in AI Text Evaluation When Author Identity is Revealed

Researchers at the University of Zurich found that large language models (LLMs) such as OpenAI’s o3-mini, Deepseek Reasoner, xAI Grok 2, and Mistral display systematic biases, but only when information about an author’s identity is provided. Without such cues, LLMs show over 90% agreement across controversial topics, debunking claims of ideological “AI nationalism.”

10 Nov 2025

Cornell Study Reveals AI’s Hidden Carbon and Water Costs and How To Fix Them

Cornell researchers have quantified AI’s rapidly growing environmental footprint, projecting that U.S. data centers could emit up to 44 million tons of CO₂ and consume over a billion cubic meters of water annually by 2030. Their Nature Sustainability study also outlines strategies that could cut these impacts by up to 86%.

10 Nov 2025

Siemens and Lehigh Develop AI-Driven Microgrids to Power Greener Data Centers

Lehigh University researchers, in partnership with Siemens, are developing autonomous microgrids to make hyperscale data centers more energy-resilient and sustainable as AI demand surges. The project integrates real-time monitoring, renewable energy, and AI-driven control systems to stabilize power and cut emissions.

10 Nov 2025

American College Of Cardiology Partners With OpenEvidence To Bring Generative AI To Cardiovascular Care

The American College of Cardiology (ACC) and OpenEvidence have announced a strategic partnership to integrate generative AI into cardiovascular care, enabling clinicians to access real-time, evidence-based insights at the point of care. The collaboration aims to enhance decision-making, improve patient outcomes, and ensure responsible AI implementation in clinical workflows.

10 Nov 2025

AI-Driven Incentive System Enhances Digital Twin Performance In 6G Networks

Researchers from China Mobile and Beijing University of Posts and Telecommunications have developed an incentive-based AI task offloading framework for digital twin networks in 6G systems. Using a Stackelberg game model and deep reinforcement learning, the system optimizes pricing and task distribution, boosting efficiency and adaptability in dynamic wireless environments.

10 Nov 2025

Mount Sinai Adopts Microsoft Dragon Copilot To Transform Clinical Workflows With AI

Mount Sinai Health System will deploy Microsoft Dragon Copilot, an AI-powered clinical assistant designed to streamline documentation, automate administrative tasks, and reduce clinician burnout. The rollout marks a key milestone in Mount Sinai’s digital transformation and commitment to responsible AI in healthcare.

10 Nov 2025

NUS Scientists Build Ultra-Thin 2D Memtransistors For Next-Generation AI Chips

Scientists at the National University of Singapore have developed ultra-thin 2D memtransistor arrays with controllable Schottky barriers, achieving unprecedented uniformity and precision. These devices demonstrated high switching ratios and over 98% image recognition accuracy, marking a breakthrough for neuromorphic and AI computing.

10 Nov 2025

AI Controls Satellite In Orbit For The First Time In Historic Space Milestone

In a world first, researchers at Julius-Maximilians-Universität Würzburg (JMU) have successfully tested an AI-based satellite attitude controller directly in orbit aboard the nanosatellite InnoCube. Using Deep Reinforcement Learning, the AI autonomously performed precise attitude maneuvers, proving that self-learning controllers can safely operate in space.

10 Nov 2025

AI Creates Ultra-Realistic Fake Photos That Fool Even Familiar Eyes

A new study shows that AI models like ChatGPT and DALL·E can now generate synthetic images of real people that are nearly impossible to distinguish from authentic photographs. Even familiar faces, including celebrities, fooled participants, underscoring deepfake realism and misinformation risks.

10 Nov 2025

AI Breakthrough Enables Rapid Antibody Design Against Emerging Viruses

Researchers at Vanderbilt University Medical Center have demonstrated that AI and protein language models can design monoclonal antibodies capable of preventing or reducing severe viral infections such as RSV and avian influenza. Their model, MAGE, can generate functional antibodies against unseen viral strains faster than traditional methods.

10 Nov 2025

AI-Powered Firefighting System Targets And Extinguishes Shipboard Oil Fires Autonomously

Researchers at Korea’s KIMM have developed the world’s first AI-driven fire suppression system capable of autonomously detecting, targeting, and extinguishing oil fires aboard naval vessels. The system uses reinforcement learning to adapt to sea motion and proved highly accurate in real-ship tests on the ROKS Ilchulbong.

10 Nov 2025

AI News

Five Minutes of Training Helps People Detect AI-Generated Fake Faces

Chiba University Researchers Develop Topology-Aware AI To Decode Brain Signals For Thought-Controlled Devices

Tufts Medical School Leads the Way in Training Future Doctors to Use AI Responsibly

Biased AI Steers Human Hiring Decisions, UW Study Finds

Penn GSE Launches First High School Curriculum to Teach Students How to Audit AI Bias

South Korean Scientists Develop AI to Detect Hidden Cracks in Structures With Unprecedented Precision

NTU and Zero Gravity Launch SG$5 Million Hub to Build Decentralised AI Systems

Scientists Use Explainable AI to Reveal How Sensors Recognize Smells

Tokyo Scientists Use AI to Decode Material Structures From X-Ray Spectra

Study Finds Hidden Bias in AI Text Evaluation When Author Identity is Revealed

Cornell Study Reveals AI’s Hidden Carbon and Water Costs and How To Fix Them

Siemens and Lehigh Develop AI-Driven Microgrids to Power Greener Data Centers

American College Of Cardiology Partners With OpenEvidence To Bring Generative AI To Cardiovascular Care

AI-Driven Incentive System Enhances Digital Twin Performance In 6G Networks

Mount Sinai Adopts Microsoft Dragon Copilot To Transform Clinical Workflows With AI

NUS Scientists Build Ultra-Thin 2D Memtransistors For Next-Generation AI Chips

AI Controls Satellite In Orbit For The First Time In Historic Space Milestone

AI Creates Ultra-Realistic Fake Photos That Fool Even Familiar Eyes

AI Breakthrough Enables Rapid Antibody Design Against Emerging Viruses

AI-Powered Firefighting System Targets And Extinguishes Shipboard Oil Fires Autonomously

Trending Stories

Latest AI News