Gal launched the RiskRubric.ai initiative within the huggingface/blog repository, creating a standardized framework for assessing AI model risk across six key areas, including transparency, reliability, and security. The work involved authoring detailed documentation and a blog post that outlined the methodology and findings, with a particular focus on the intersection of security and safety. Gal updated the blog’s configuration using Markdown and YAML, integrating new visual assets and linking to relevant Hugging Face datasets. The project demonstrated depth in AI safety and technical writing, delivering a maintainable content management solution that supports ongoing risk assessment and clear communication of results.