As we're ramping up DALL-E access, safe usage of the platform is our highest priority. Our filters aims to detect generated text that could be sensitive or unsafe. We've built the filter to err on the side of caution, so, occasionally, innocent prompts will be flagged as unsafe.

Although suspensions are automatic, we manually review suspensions to determine whether or not it was justified. If it wasn’t justified, we reinstate access right away.

If you have any questions on your usage, please see our Content Policy.

Did this answer your question?