Dear OpenAI Team,
I’m writing to express serious frustration with the current state of your image generation tools—specifically the recent “upgrades” that have crippled functionality for creators like me.
Here’s what I’ve experienced:
What’s Broken
-
Face Referencing Is Inconsistent
Despite uploading a face reference image, the tool fails to reproduce consistent likeness—even when using the exact same prompt. This is unacceptable for character-driven projects.
-
Rate Limits Are Overly Harsh
Trying to iterate on a visual—adjusting poses, muscles, or clothing—now hits cooldowns constantly. I can’t even refine a single design without being locked out for 10+ minutes.
-
Prompts Are Getting Flagged for No Reason
Clean, anatomically focused prompts (like a side view squat in gym wear) are being blocked with vague “content policy” violations, even when nothing inappropriate is requested.
-
Body Proportions Are Wildly Inaccurate
Female characters end up with bulky, masculine arms, or glitched proportions, even when detailed measurements are provided. This wrecks the consistency of carefully developed designs.
-
Old Tools Worked Better
Before the so-called upgrade, I could at least generate my character consistently, with face accuracy and fewer restrictions. Now I get more broken results, fewer options, and more wait time.
What I Need as a Creator
- Functional face transfer from reference
- Control over muscle tone and proportions
- The ability to create neck-down character shots
- Fewer rate limits during iterative sessions
- A system that helps, not hinders, storytelling
If OpenAI wants to support real creators—those building deep lore, stories, and characters—you need to fix the system you broke and start empowering creators, not stifling them.
Right now, it feels like you’re tuning the platform for casual toy use, not for professionals or world-builders.
I hope this gets taken seriously, because this platform has incredible potential. But if these issues aren’t addressed soon, you’re going to lose the people who care the most.
Sincerely,
A seriously disappointed creator
2 Likes
I feel the same 
My text I wrote w days ago never send because I thought IAM alone with this …
As a dedicated user and supporter of OpenAI’s tools, I want to voice something that many in the community feel but few say loud enough: The ever-increasing restrictions and overly aggressive filtering systems are suffocating innovation and user expression.
Yes, I understand the importance of safety, copyright, and ethical boundaries. But there’s a fine line between protecting users and limiting them. When policies start blocking not only harmful misuse but also legitimate, creative, and original ideas, we’ve lost the plot.
The rise of AI has sparked a creative renaissance. People are building, dreaming, experimenting. But instead of empowering this wave, we’re met with red tape, blocked prompts, and a constant feeling of being monitored and limited – even when we’re clearly not violating any rules.
Of course there will always be bad actors. But punishing the majority for the actions of a few is the fastest way to drive away your loyal, paying user base.
We don’t need less safety – we need smarter safety. Context-aware, user-trusted systems that differentiate between abuse and creativity.
If OpenAI wants to stay at the forefront of this revolution, it must stop treating its community like potential threats, and start treating us like partners in progress.
to be completely honest i dont care about safety. I’m adult. I pay for the services, and everything. therefore I shouldn’t feel like a child when I’m creating stuff. If openai wants to charge more or do something to prove that I’m adult so i can have more freedom to create then so be it. but the trash guidelines are just that trash. and now. even more limits. like what am I paying for anymore??? more restrictions. Less freedoms, half ass upgrades. like are you guys begging for your viewer base to go some where else. because thats what it feels like to me. Im tired of it telling me it knows how i feels but it doesn’t. so i wish it would just shut up on that. and when I ask it to do the most basic things. it messes it up and i have to keep asking it not be to be lazy with no exaggerations. 12 times. in a row. and then my favorite. I will ask chat to do a image and of course its not right so i ask it to do it again. and I kid you not. 3 times in a row. i get the same damn image. (the exact one) 3 times even with i get clear directions on what I want etc.
and then its tell me oh Im sorry I will fix it next time. Like why is there even a chance of it needing to do it “right” next time. when the directions were clear the first time. and its not like im asking it to do something that it might not be able to do. and my most favorite is when I catch it in lies. telling me it cant do this or than. and a few minutes in another chat window. I was doing the exact thing. so again. Thank Openai. for your “upgrades” that keep breaking everything. and making it alot less enjoyable. if this keep happening I have no problem canceling my sub. and going to some where, where I dont want to throat punch the damn thing every day. maybe Grok 3 or something. Im getting to the point where I dont care im going to have to transfer over 300 pages with of data to somewhere else that actually works. no stress, no drama, no retarded guidelines. im sorry that you and your censors are hyper-sensitive. as a result, ive lost days working because of that. then I ask chat why. and it can’t tell me why. just that whatever I was doing triggered the ai to have a fit.
i could understand If I was trying to make porn on your platfrom. Im not there are way better ai apps for that. but yeah being treated like a child for something I paid for is non-sense. if you are really that worried about it. give us a way to prove we are adults and not children. if that is your only reason. or create an payment thing where you charge like 30 dollars a month with less restrictions So i dont feel like a child. Eitherway if this non-sense keeps up then ya. I’m 100 percent confident that alot of people will start to leave. and i hope they keep leaving until openai gets their shit together otherwise. im okay with them going under.
1 Like
Probably the same AI model they used to cancel all those “Dead” SS accounts, or remove references to women from those military websites. If you put your faith into a jumped up spell-checker, expect to be disappointed.