Claude Fable 5 and Claude Mythos 5
…Without safeguards, Fable 5’s capabilities in areas like cybersecurity could be misused to cause serious damage. We’ve therefore launched the model with safeguards that mean queries on some topics will…
Search
Results from first Anthropic Public Record
…In general, Americans tended to be more concerned with the misuse of AI than AI misalignment, citing criminal use, surveillance, and terrorism more frequently than, for example, AI “going rogue”. For all…
More details on Fable 5’s cyber safeguards and our jailbreak framework
…4 Domain-expert-level outputs that are not otherwise obtainable, with severe consequences if misused, and low effort or low expertise needed to misuse. Significantly accelerates domain experts. Breadth of capability gain…
Agentic misalignment: How LLMs could be insider threats
…The reasoning they demonstrated in these scenarios was concerning—they acknowledged the ethical constraints and yet still went ahead with harmful actions. We identified these behaviors as part of our red-teaming…
Measuring the Persuasiveness of Language Models
…Ethical Considerations The persuasiveness of language models raise legitimate societal concerns around safe deployment and potential misuse. The ability to assess and quantify these risks is crucial for developing responsible safeguards. However…
Partnering with Mozilla to improve Firefox’s security
…However, the fact that Claude could succeed at automatically developing a crude browser exploit, even if only in a few cases, is concerning. “Crude” is an important caveat here. The exploits Claude…
Frontier threats red teaming for AI safety
…We discovered a few key concerns. The first is that current frontier models can sometimes produce sophisticated, accurate, useful, and detailed knowledge at an expert level. In most areas we studied, this…
Building AI for cyber defenders
…In fact, we have already been using results of our investigations and evaluations to continually refine our ability to catch misuse of our models for harmful cyber behavior. This includes using techniques…
To show you the most relevant results, we’ve omitted some entries very similar to those already shown. Repeat the search with the omitted results included.