Anthropic Publishes New Claude AI Constitution

Anthropic released a new constitution detailing the values and training principles that guide its Claude AI models, aiming to improve safety, ethics and transparency.

• Sets core priorities: safety, ethics, guideline compliance, helpfulness
• Used directly in training, including synthetic data and response ranking
• Shifts from rigid rules to principle-based reasoning
• Includes hard constraints such as bans on bioweapons assistance
• Released publicly under a Creative Commons CC0 1.0 license

To view the source of this information, click here

To contact the reporter on this story:
Georgi Azar in New York at gazar4@bloomberg.net

To contact the editor responsible for this story:
Ilya Banares ...