Guides · Updated 27 October 2026

Will my video get flagged as sensitive on X?

Videos get sensitive-content warnings on X just like photos do, and a single frame can be enough. Here's how video labeling works, what tends to trigger it, and how to check a clip before you post it.

Can a video be marked sensitive on X?

Yes. X's Adult Content Policy and Violent Content Policy cover "images and videos" alike. If a video contains adult content or graphic media, X expects it to carry a content warning. If it doesn't, it can be labeled by X's systems or after reports. A labeled video shows up blurred with a warning and a Show button, and X's policies say viewers under 18 or without a birth date on their profile can't open it.

Live video is stricter. X's policies say adult and violent media isn't allowed in live video at all, and a warning doesn't make it acceptable there.

Why videos get flagged more often than you'd expect

• It only takes a moment. A video is a long string of images. Two seconds of a fight scene, a wound, or someone changing clothes in the background can be enough, even if the rest of the clip is harmless.
• Gaming and movie clips. Realistic blood and gore in a game or horror movie can look like graphic media to an image model. X's violent content policy allows "painted or illustrated art," but a frame-by-frame system can't always tell a realistic game from real footage.
• Fitness, dance and swimwear. Lots of skin in motion can be mistaken for adult content, especially in low light or with skin-toned clothing.
• Medical and beauty content. Surgery, injections, tattoos and wound care can look like "serious physical harm" to a classifier. X's policy lists medical procedures shown "in graphic detail" as needing a warning, while scars, stitches and healed wounds are allowed.
• Reaction and news clips. Accident, crime or protest footage is a common source of graphic-media labels.

How X checks videos (and what it doesn't publish)

X doesn't publish exactly how it checks videos, which frames it looks at, or its thresholds. What we do know is that in 2023 Twitter open-sourced the training code for an image classifier called pNSFWMedia in its trust and safety models. It uses a pretrained image model to turn an image into numbers and a small neural network to turn those into a sensitivity score. A common way to apply an image model to video is to sample frames across the clip and look at the highest-scoring ones. That's also what Will It Flag? does.

How to check a video before you post it

1. Open Will It Flag? and choose the Video tab.
2. Pick your clip. Frames are sampled across the whole video and checked in your browser. The video isn't uploaded anywhere.
3. Look at the timeline. It shows which moments scored highest, so you know exactly which part to trim, blur or cut.
4. Edit and check again, or decide to add a one-time content warning yourself.

The result is an estimate based on the same model design X published. It isn't X's actual decision, and X also uses context, reports and human review. Treat a high score as "this part might get a warning," not as a guarantee either way.

Tips to avoid an unnecessary label

• Trim the risky seconds. If one scene scores high, cutting it is usually the simplest fix.
• Choose a clean opening. The first frames make the first impression in the feed, so avoid opening on gore or a lot of skin.
• Label it yourself when it's genuinely graphic. For real news or medical footage, add a one-time warning. X's policies ask you to, and it protects your account from being labeled.
• Keep your profile clean. Graphic or adult clips should never be used in a profile photo, header or live video.

Will It Flag? is an independent tool and is not affiliated with X Corp. X's rules and app menus change from time to time; this guide reflects X's published Help Center policies as of 27 October 2026. When in doubt, X's own Adult Content and Violent Content policies are the source of truth.

Frequently asked questions