I’m running into what looks like a false-positive safety filter issue with Gemini Omni video editing.
When I upload a normal video and ask for a harmless edit, the generation instantly fails before it even starts. Example prompt:
“apply a green cinematic color grade to the uploaded video while preserving the same scene motion and content”
This gets rejected as harmful content in Flow, and I’m also seeing the same type of instant rejection in the Gemini app. The prompt is only asking for a color grade / lighting change, not anything unsafe.
This seems related to the previously mentioned bug b/515000564 where simple video editing prompts were falsely triggering filters. Is anyone else still seeing this? Has there been any update on whether this issue is fixed or still being investigated?
I absolutely HATE this update. Prompts that I’ve used before are now being reflected or “I can’t generate that video. Try describing another idea. You can also get tips for how to write prompts and review our video policy guidelines.” . Then I try using FLO and it completely disregards the image uploaded and does its own thing. Who is this update for?
Same issue here. A video of me, taken on my phone, with the request to make me wear an astronaut suit gets rejected as “against our policies”. IDK why they’re advertising these capabilities when we can’t use them.
Hello Everyone,
If you are experiencing this issue within the Gemini application, I suggest filing a report directly through the Help section of the Gemini App. Navigate to Settings & help (gear icon in the bottom left) > Send Feedback , and follow the prompts to submit the request with the relevant details to ensure appropriate support. Note that this forum is specifically for Google AI Studio and Gemini API inquiries.