I am a paying ChatGPT Plus subscriber and I’m writing to express my serious frustration with the current state of your image generation policies. Over the past two days, I have been repeatedly blocked from generating a simple, educational image for my fitness brand.
Despite multiple careful attempts to reword and adapt the prompts — including following advice directly from your system (ChatGPT) — the requests are still blocked.
These are not NSFW images, nor are they suggestive, offensive, or inappropriate in any way.
They are meant to teach proper exercise form. And yet, I’m prevented from doing something as basic as adding a posture annotation near the spine or placing a logo on a sports top — which is standard in the fitness industry.
This is unacceptable.
- I am paying for a service that is actively blocking me from completing my work
- I have wasted over two full working days trying to get one simple image approved
- The moderation system shows zero awareness of context or intent
- There is no transparency or explanation of what triggers these blocks
–
All of these elements are standard in thousands of fitness illustrations across the world.
Despite this, your system has blocked my image generation attempts again and again, regardless of how careful, neutral, or professional the prompts were. Even minor changes to annotation wording, arrow placement, logo positioning or background tone have triggered moderation flags.
–
So far, I’ve made over 95 prompt attempts and visual adjustments and counting , including:
- 50+ generation prompts and reworded instructions
- 20+ annotation and label changes
- 10+ layout/style corrections (color, text position, branding)
- 15+ full image blocks due to your content filter
- Ongoing time lost trying to get a single asset built
And the image is still not complete.
All of this is for a single visual, which remains entirely appropriate, instructional, and within the bounds of professional educational content.
–
Your system appears to treat:
- Fitness poses
- Labels near the spine, chest, or hips
- Any visual emphasis on body positioning
…as potential violations — even when the prompt context is clear, safe, and aligned with educational standards.
The filters are so sensitive that they actively block legitimate professional usage, and offer no meaningful explanation or override path.
–
This has cost me hours of work, broken my creative flow, and put my paid subscription into question.
I am not asking for anything unreasonable — only that your system recognizes legitimate, professional use cases and allows creators like me to do our work.
I truly hope OpenAI chooses to stand with creators, not stand in their way.
–
Please treat this as a formal complaint and escalate it accordingly.