How long must the EU AI label stay on screen? Cue timing

The Commission gives no number of seconds. Its icons page says that if a disclosure appears for a limited time, it should remain visible long enough to be read and understood by users with cognitive or processing difficulties. So the safe design question is how long the text is on screen, not a fixed minimum. This is not legal advice.

Quoted from the Commission icons page (last updated 24 September 2026), read 2026-10-01. Timing fields are from Video captions.

What does the Commission page say about timing?

Two nearby rules matter. The icon should be clearly perceivable at the latest at first exposure, and the accessibility advice for a limited-time disclosure is the one above. The simplest way to satisfy both is to show the disclosure from the first frame and keep it for the whole clip.

How do I set start and end times for a disclosure?

Standalone captions accept cues (or segments), each with text, start and end in seconds. The docs say this authored overlay skips speech-to-text, so it works on silent clips too. script_text, words, cues and segments are mutually exclusive, so a disclosure cue cannot be mixed with a speech-aligned script in the same request.

{
  "video_url": "https://media.sume.com/artifacts/artf_demo/clip.mp4",
  "cues": [
    { "text": "AI-generated", "start": 0, "end": 12 }
  ]
}

How does clip length change the choice?

Avatar video scripts are accepted at 4-60 seconds estimated duration, inclusive. Short clips make a full-length cue easy; the end value can match the clip length. For a longer cut, video trim takes one workspace clip and a range, so you can trim first and set the cue to the new length.

Where should the cue sit in the frame?

Timing is only half of it. Keep the cue clear of other overlay elements; see keep captions off the label and burning the label into the first frame.