AI CONTENT LABEL CHECK

用了 AI,不等于只补一个标签就够了。

适用于 AI 图片、AI 视频、AI 配音、数字人、换脸、拟声、虚构演绎与 AI 辅助改写。审核会分清“哪里使用了 AI”“是否容易让人误认”“还缺什么权利或事实说明”。

一句话怎么用

使用 $media-publish-check 审核我准备发布的 AI 内容。
AI 用在:封面图、视频中的数字人和配音;故事由我本人创作。
我要发小红书和抖音。请检查需要如实说明的部分、素材/肖像/声音权利,
并分别给两个平台的发布页标注建议。先只给结论、最多 3 个问题和修改稿。

会检查什么

AI 的实际使用范围

区分只用 AI 做封面、局部修图、AI 配音、全片生成、数字人或虚构演绎;不把不同情况混成同一种风险。

如实披露

检查视频内、封面、正文与发布页是否需要清楚说明;提示缺少的事实,而不是用含糊话术规避理解。

人物与素材权利

核对真人肖像、声音、商标、照片、视频片段、训练/生成素材和授权范围;“标注 AI”不能自动解决侵权。

平台版本

分别检查要发往的平台及其发布场景;同一内容不会被假设为适合每一个平台。

关键边界:即使只有封面由 AI 生成,也应如实检查和选择适用的 AI 标注。具体发布页选项和平台规则会变化,最终以发布时页面和官方规则为准。
AI CONTENT LABEL CHECK

Using AI is not solved by adding one generic label.

Use this for AI images, video, voice, avatars, face or voice imitation, reenactments, and AI-assisted rewrites. The review separates how AI was used, what could mislead viewers, and what rights or factual disclosures remain.

Use it in one prompt

Use $media-publish-check to review my AI-assisted content.
AI was used for the cover image, a video avatar, and voiceover; the story is mine.
I will publish on Xiaohongshu and Douyin. Check truthful disclosures and image/voice/asset rights,
and give platform-specific label guidance. Return only a decision, up to 3 issues, and revisions.

What it checks

Actual AI use

Distinguishes AI covers, retouching, voice, fully generated video, avatars, and reenactments instead of treating every use as identical.

Truthful disclosure

Reviews whether the video, cover, post body, and publishing page explain material AI use clearly rather than obscuring it.

People and asset rights

Checks likenesses, voices, brands, photos, clips, generated assets, and authorization scope. An AI label does not cure infringement.

Platform versioning

Reviews each intended platform and its publishing context; one version is not assumed fit for every platform.

Important boundary: Even when AI is only used for a cover, disclose it truthfully where required. Publishing-page options and rules can change; confirm them at publication time.