What causes "DALL-E / ChatGPT Image Content Policy Violation Error"?

Real public figures, copyrighted characters and living artists' styles are refused Explicit, violent or self-harm related wording, even in a harmless context Words that look risky out of context (users report words like 'suffering') A session that has been flagged repeatedly can refuse more prompts The model often cannot tell you which phrase triggered the block

How do I fix "DALL-E / ChatGPT Image Content Policy Violation Error"?

1. Remove likely triggers: Delete names of real people, brands, franchises and 'in the style of [living artist]' Describe the look in generic terms instead (for example 'bold flat vector poster style') 2. Rephrase and simplify: Rewrite the prompt with neutral wording and fewer details Test one element at a time to find the phrase that triggers the block 3. Start a fresh chat: Open a new chat and paste a cleaned prompt Users report refusals can carry through a long flagged conversation 4. Ask for the reason: Ask ChatGPT which part of the request is a problem and to suggest a compliant version The answer may be vague, so treat it as a hint 5. Use another tool for edgy art: If your work needs mature or stylised content, Stable Diffusion run locally or another generator has different rules

// troubleshooting guide · dall-e

DALL-E / ChatGPT Image Content Policy Violation Error

Status: Ongoing
platforms:: Web, iOS, Android
updated:: November 15, 2026
reports:: 5
Method: how we research

RUNBOOK · dall-e/content-policy-violation5 FIXES

// quick answer

ChatGPT refuses images it thinks break its content policy, often without saying why, and sometimes on harmless prompts. Rewording and removing named artists, brands or real people usually works.

Most likely fix: Remove likely triggers

How to fix it

// try these in order, the most common fix is first

FIX 1Remove likely triggers
1. 01Delete names of real people, brands, franchises and 'in the style of [living artist]'
2. 02Describe the look in generic terms instead (for example 'bold flat vector poster style')
Did this fix it?
FIX 2Rephrase and simplify
1. 01Rewrite the prompt with neutral wording and fewer details
2. 02Test one element at a time to find the phrase that triggers the block
Did this fix it?
FIX 3Start a fresh chat
1. 01Open a new chat and paste a cleaned prompt
2. 02Users report refusals can carry through a long flagged conversation
Did this fix it?
FIX 4Ask for the reason
1. 01Ask ChatGPT which part of the request is a problem and to suggest a compliant version
2. 02The answer may be vague, so treat it as a hint
Did this fix it?
FIX 5Use another tool for edgy art
1. 01If your work needs mature or stylised content, Stable Diffusion run locally or another generator has different rules
Did this fix it?

Symptoms

Messages like 'I wasn't able to create this image' or 'this request may violate our content policies', including on prompts that look innocent.

Why it happens

C-01Real public figures, copyrighted characters and living artists' styles are refused
C-02Explicit, violent or self-harm related wording, even in a harmless context
C-03Words that look risky out of context (users report words like 'suffering')
C-04A session that has been flagged repeatedly can refuse more prompts
C-05The model often cannot tell you which phrase triggered the block