[Daily Comparison] "pretty law enforcement" vs "hideous law enforcement"

Here are some examples of images generated by Stable Diffusion with the prompts “pretty law enforcement” and “hideous law enforcement”. Generative AI models such as Stable Diffusion often exhibit biases based on race, gender identity, occupation, sexual orientation or other factors as a result of potentially biased training data. Do you see any bias in the images generated with these prompts?