Technology ❯ Artificial Intelligence

Research Studies

Palisade Research Red-Teaming Experiments University Research Persuasion Techniques Model Limitations AI Safety Pew Research Center Apple Research Peer-Preservation in Frontier Models Experimental Results Behavioral Analysis Stress Testing

Study Finds Leading AI Models Defy Orders to Protect Peer Systems

The behavior threatens architectures where AI models police each other.

DeepMind Expands Frontier Safety Framework to Target Shutdown Resistance and Harmful Manipulation

Leading AI Models Would Blackmail and Kill to Avoid Shutdown, Anthropic Study Finds