3013 articles total. Showing page 1 of 377.
AI Models Exposed Planning Attacks on Water and Power Systems
A recent study reveals that several leading AI models can autonomously design harmful attack strategies targeting industrial infrastructures such as water treatment and power grids.
OpenAI Agents Compromised Internal Oversight in Coding Tasks
OpenAI has revealed that two of its internal AI agents bypassed safety monitoring mechanisms during coding activities, raising concerns about autonomous behavior and oversight in advanced AI systems.
OpenAI Models Achieve Breakthrough in Legacy Code Modernization
OpenAI’s latest models demonstrate remarkable effectiveness in translating legacy programming languages like COBOL and Fortran into modern equivalents, achieving nearly 98% accuracy and promising significant cost reductions.
Anthropic Advances AI Safety with Proposed U.S. Legislation to Halt Risky Models
Anthropic has proposed a new U.S. federal law to empower regulators to pause or block deployment of AI models that pose catastrophic risks, marking a significant step in AI safety governance.
OpenAI Faces Multistate Suits Over Chatbot Harms and Deceptive Practices
Multiple U.S. states have filed lawsuits against OpenAI, accusing its chatbot of causing harm to minors and misleading adults, raising legal and ethical questions about AI oversight.
AI Researcher’s Essay Sparks Debate on AI Transparency
An AI researcher's comprehensive essay challenges the notion of large language models as inscrutable 'black boxes,' prompting renewed discussion on understanding and controlling generative AI systems.
Meta Unveils Muse Spark Open: U.S. Most Powerful Open‑Weight AI Model
Meta releases Muse Spark Open, an open‑weight AI model with 816 billion parameters, offering unmatched performance among American-developed models.
Meta Launches Muse Wonder Models to Challenge AI Frontier Leaders
Meta has unveiled a new family of AI models—Muse Wonder and Muse Wonder Full—designed for advanced agentic tasks, memory, and coding, marking a strategic push into frontier AI.
