Thinking about standards for AI model validation and reporting