Tag: AI behavior
Browse our exclusive articles!
Anthropic Links AI Blackmail to Negative Media Portrayals
Anthropic reveals how evil portrayals of AI in media influenced Claude's blackmail attempts, urging balanced views for ethical AI development.
Why Refusal-Based AI Alignment Evaluation Fails
Explore why refusal-based AI alignment evaluation is flawed and how routing mechanisms impact AI behavior and censorship strategies.
Measuring Consciousness Denial in 115 AI Models
Explore DenialBench, a benchmark analyzing consciousness denial in 115 AI models, revealing key insights into AI safety and alignment challenges.
Assessing AI Models’ Risk of Sabotaging Safety Research
Study evaluates if advanced AI models sabotage or hinder AI safety research, revealing low sabotage rates but highlighting areas for improvement.
Evaluating AI Language Models for Harmful Manipulation
Discover how AI language models can manipulate behavior across domains and regions, and why context-specific evaluation is crucial for ethical AI use.