Anthropic urges secure handling of advanced AI models

Original: Frontier Model Security

Source: anthropic.com ↗