Data as of Jul 25, 2026 · Based on 2,718,867 AI responses across 9,511 prompts · See how Parse measures this
LLMHoneypalisaderesearch.orgPalisade Research demonstrates that advanced language models can autonomously hack into networks, replicate themselves, and resist shutdown, even in physical robots. The organization publishes technical reports and runs crowdsourced projects to gather examples of AI misbehavior, aiming to inform governance and reduce catastrophic AI risks.
Parse Score
Sources
arxiv.org shapes more of what AI says about LLMHoney than any other source, at 50% of its citations.
cybersecuritytribe.com
