"Quality issues" rejection on AI-generated photo — what specifically is wrong?

Hi everyone,

New contributor here, hoping for experienced eyes on a rejection. This image was refused for "Quality issues" — it's AI-generated (labeled as such on submission, Photos category, no people):

Traditional Pakistani truck art decorated lorry at sunset with vibrant painted patterns

Technical details:

• Generated at 1920×1280, upscaled 2× to 3840×2560 to meet the 4MP minimum, submitted as JPEG
• The rejection lists the usual suspects (exposure, soft focus, filtering/artifacts/noise) but doesn't say which one

Before I rework my pipeline, I'd like to know what I'm actually fixing:

1. Is this upscaling softness? (I used a neural upscaler rather than Lanczos, but 2× is 2×.)
2. Or AI-generation artifacts somewhere I'm not seeing at normal zoom?
3. Or is the busy composition/detail level itself the problem?

Happy to share 100% crops of any area if that helps diagnose. Thanks in advance!