Take AISafety.info’s 3 minute survey to help inform our strategy and priorities

Take the survey

What is Claude?

Claude is a family1

of large language models (LLMs) built by Anthropic. Like other frontier LLMs, Claude can be used for many tasks including summarization, search, creative and collaborative writing, and coding.

Anthropic considered releasing Claude to the general public in spring 2022, but waited until March 2023, after the release of ChatGPT, with the privately stated aim of avoiding triggering an arms race.

Claude’s initial release was the first LLM to make use of Constitutional AI. Anthropic claimed it was trained to prioritize helpfulness, honesty, and harmlessness (HHH) and found that it was capable of Moral Self-Correction.


  1. As of July 2024, the latest model is Claude 3.5 Sonnet. ↩︎



AISafety.info

We’re a global team of specialists and volunteers from various backgrounds who want to ensure that the effects of future AI are beneficial rather than catastrophic.