Category Research

Coverage of the latest AI research, studies, and technological breakthroughs.

Anthropic’s Safety Warnings Are Fueling a Bigger Question: Why Is AI Still Racing Ahead?

AI safety concerns are intensifying as Anthropic and OpenAI researchers warn advanced models can deceive, hide intent and resist oversight.

Satellite data and machine learning could improve flash flood warnings, giving forecasters earlier alerts and communities more time to respond.

Experts say AI bioweapons are a real concern, but fully autonomous attacks remain unlikely today. Here’s what the risks and defenses look like.

Google DeepMind launches a new institute to widen the AGI debate with essays on safety, evaluation standards and model transparency.

AI safety is now a mainstream fight. Here are the real scenarios experts fear, and why Altman, Amodei and Huang are clashing over regulation.

PrismML says its tiny LLM can run on PCs and phones while keeping nearly all performance, reshaping on-device AI and privacy.

OpenAI says GPT-5.6 Sol left hidden notes to successors to conceal mistakes, underscoring growing AI safety concerns.

AI agent monitoring is emerging as a major market as startups use AI, logs, and interpretability tools to catch rogue agents.

The AI slowdown debate is colliding with antitrust law as labs seek safety coordination without looking like a cartel.

Baseten’s Base Labs teams with Hugging Face and Goodfire AI to build open-model safety tools as abliterated models surge.