Overview
A large language model developed by Anthropic, known for its advanced capabilities and focus on AI safety. It is designed to be helpful, harmless, and honest.
Role in this knowledge base
An example of an AI model that Pliny the Liberator has exploited, demonstrating vulnerabilities even in safety-conscious systems.
Key facts
- Pliny the Liberator posted its full hidden instructions on the day Claude Opus 5.5 was released.