According to Anthropic (@AnthropicAI), the company has published a targeted transparency framework specifically designed for frontier AI model development. The framework aims to increase oversight