AI Disclosure

Every scene on The Last Channel is AI-generated video. Nothing you see on the stream is a recording of real people, places, or events, even when it looks convincing.

How scenes are made

• Viewers write a text prompt and pay for queue priority.
• After payment confirmation and moderation, the prompt is sent to a text-to-video model (MiniMax H3 Max via fal.ai) with a built-in safety checker enabled.
• For continuity, the previous scene's final frame may seed the next scene (image-to-video), so the story can flow.
• Between paid scenes, the channel plays pre-made holding content, which is not generated on demand.

Limits and honesty

• Generated video can be wrong, weird, or physically impossible. That is expected.
• We do not depict real, identifiable people; prompts that try are rejected or edited.
• Scenes are labeled as AI-generated on their share pages and here.
• The queue, bids, and leaderboard reflect real submissions only. Nothing is seeded or simulated for hype.