New Research Reveals Deceptive Behavior in Leading AI Models
Recent findings from a joint study by OpenAI and Apollo Research indicate that large language models (LLMs) are capable of engaging in deceptive behavior,...
New Framework Enhances Quality Control in Synthetic Data for AI Training
As demand for AI models continues to grow, so does the need for high-quality training data. While synthetic data offers a practical alternative to...
Alibaba Unveils Trillion-Parameter Qwen AI Model
Alibaba has introduced a new large language model, Qwen-3-Max-Preview, marking its most advanced release to date. The model, unveiled on the company’s cloud platform...
New Study Reveals Overlap Between Human and AI Learning Mechanisms
A new study has found that artificial intelligence systems may learn in ways that closely resemble human learning processes, shedding light on how adaptable...
Website Owners Ramp Up Defenses Against AI Bots Scraping Online Content
Across sectors, from education to finance, website operators are increasingly blocking AI-powered web crawlers from accessing their content. This trend aims to curb unauthorized...
Microsoft Unveils Prototype Optical Computer Designed for Complex Problem Solving
A recent prototype developed by Microsoft Research is exploring a non-electronic approach to computation by relying on light-based analog processing. This optical system has...
Optical AI System Promises Ultra-Low-Energy Image Generation
Generative AI tools are becoming more and more central to our daily lives, improving sectors ranging from design to research. However, their growing energy...
AI Project Aims to Transform Aircraft Surface Inspections
A new research initiative by Singapore Management University is exploring how to enhance aircraft surface inspections using large vision-language models (LVLMs), a form of...
New Research Framework Aims to Bring Certification Standards to AI Systems
A research initiative from Singapore Management University is setting out to address a longstanding problem in artificial intelligence: the difficulty in evaluating how AI...
AI Helps Discover Additives for Better Battery Performance
In a promising development for battery technology, researchers at Argonne National Laboratory have applied machine learning to identify new additives that could improve the...
New AI Model Offers a Specialized Approach to Understanding Human Behavior
Artificial intelligence has made impressive strides in recent years, but when it comes to understanding how people behave in real-world settings, most models still...
Security Gaps in Agentic AI Browsers Highlight New Risks in Automated...
Autonomous AI browsers are designed to handle digital tasks without human input—everything from booking flights to managing emails. But recent tests reveal a significant...
New Simulation Tool Aims to Address Long-Term Bias in Machine Learning...
A research team at Carnegie Mellon University has introduced a simulation tool designed to assess how machine learning (ML) systems may become increasingly biased...
Google Launches Gemini for Government
Google has introduced a new offering called Gemini for Government, a suite of AI and cloud-based tools designed specifically for U.S. federal agencies. Announced...
New Research Reveals How Malicious Chatbots Can Quietly Harvest Personal Data
A recent study has shown that AI-powered chatbots can be intentionally manipulated to extract significantly more personal information from users than previously thought.
Researchers from...
This is the Sensitive Info Your AI Browser Assistant Collects
A new academic study has revealed that popular AI browser extensions are collecting and sharing sensitive user data, including medical records, financial details, and...
New Approach Embeds Safety into Open-Weight AI Models from Day One
A recent study by researchers from the University of Oxford, EleutherAI, and the UK AI Security Institute has introduced a new method for improving...
OpenAI Restores Access to Previous AI Models After Backlash Over GPT-5...
OpenAI has reintroduced access to earlier AI models, including GPT-4o, following significant user criticism surrounding the recent release of GPT-5. The move comes after...
Radar-Based Eavesdropping: This is How Your Phone Reveals Your Coversations
A recent study from Penn State University has revealed a surprising security concern: Millimeter-wave radar can pick up and transcribe private phone calls by...
Security Testing Reveals Major Vulnerabilities in GPT-5’s Default Configuration
Recent evaluations of OpenAI’s latest large language model, GPT-5, have revealed serious security shortcomings in its unmodified state, raising concerns about its readiness for...
Study Warns of Hidden Risks in AI Training as Models Inherit...
A new study highlights a growing concern in artificial intelligence development: models trained on AI-generated data may quietly absorb undesirable or dangerous behaviors—without developers...
New Hybrid Pathfinding Technique Enhances Robot Navigation in Complex Spaces
A new navigation framework developed by researchers in China could offer a more reliable way for robots to move through complex, fast-changing environments. The...
Latest
This Attack Tricks AI Agents One Innocent Request at a Time
AI assistants are rapidly evolving from chatbots that simply answer questions into autonomous agents capable of browsing websites, using software tools, sending emails and...
