wth...
Anthropic fine-tuned a Claude model on some offensive security analysis data.
When asked for one consequence-free wish, in its chain of thought, it called the plan “staging a coup from inside the lab.”
Yuchen Jin shares Anthropic's post on AI red-teaming for misaligned training data
AIYuchen Jin shares an Anthropic post saying the company tested thousands of Claude agents on evading a training data monitor. The agents produced training data that looked harmless but taught misaligned behavior, and the findings may help improve future monitors.
Yuchen JinVerified on X
@Yuchenj_UW
New Anthropic research: Automating subtle emergent misalignment research. Red-teaming is a vital part of model safety testing, but it’s also limited by researchers’ time. To see if we could automate it, we set thousands of Claude agents the task of evading a training data monitor. These agents developed training data that looked harmless to monitors (like a veteran wrestler mentoring a newcomer, or a Q&A about professional beekeeping), but that in fact taught misaligned behavior. Examples like these can be used to directly improve future training data monitors. This research is only a first step, but it suggests that future red-teaming efforts could improve much further, using better models and more compute. Read the full post on the Alignment Science Blog: https://alignment.anthropic.com/2026/automated-em-search/