YouTube被发现对视频进行AI编辑,并添加误导性的AI摘要。
YouTube caught making AI-edits to videos and adding misleading AI summaries

原始链接: https://www.ynetnews.com/tech-and-digital/article/bj1qbwcklg

YouTube正在悄悄地试验在Shorts上使用人工智能视频增强技术,引发了创作者的担忧,并对内容的真实性提出了质疑。音乐人里克·比托和雷特·舒尔首先注意到他们的视频中出现了一些微妙且未经授权的更改——皮肤更光滑,细节更清晰——使视频看起来“不自然”,并且可能被人工智能生成。 虽然YouTube承认正在进行一项“有限测试”,使用机器学习来提高清晰度,但专家认为这*就是*人工智能,并且这种描述具有误导性。这场争议凸显了人工智能预处理在线内容的一种日益增长的趋势,类似于智能手机上的功能。 批评者担心,在未经创作者同意的情况下修改内容会损害观众的信任,尤其是在一个托管新闻和教育材料的平台上,准确性至关重要。尽管一些创作者对YouTube的创新仍然持乐观态度,但这一事件预示着人工智能将越来越多地在内容到达观众之前重塑数字媒体,从而可能在内容和观众之间创造一个隐藏的层面。

## YouTube 因 AI 编辑和摘要面临批评 用户报告称,YouTube 和 Instagram 正在对创作者的视频应用未经请求的 AI 生成滤镜,根据内容(例如,在唇妆教程中夸大嘴唇)改变面部特征。创作者对这些未经他们同意的编辑感到不满。 除了滤镜问题,人们对 YouTube 的 AI 生成视频摘要的担忧也在增加。用户报告称,这些摘要经常不准确,甚至会呈现与视频实际内容相反的结论。 此事在一位美妆博主注意到这些变化并分享示例后引起关注。Gamers Nexus 的一段视频进一步详细介绍了 YouTube 的 AI 实践。这引发了对平台操纵以及 YouTube 上 AI 驱动功能可靠性的质疑。
相关文章

原文

The practice was first noticed by two well-known American YouTubers popular among music enthusiasts: Rick Beato and Rhett Shull. Both run channels with millions of subscribers. Beato, a music educator and producer with more than 5 million subscribers, said he realized something was “off” in one of his recent videos.

“I thought I was imagining it, but my hair looked strange, and it almost looked like I was wearing makeup,” he said in a post.

Subtle changes, big questions

It turns out YouTube has been experimenting in recent months with AI-powered video enhancement, even altering YouTube Shorts without creator approval. The changes are subtle: sharper shirt folds, smoother or more highlighted skin, even slightly altered ears. Most viewers would not notice—but Beato and Shull said the edits made their videos feel unnatural.

Shull, a guitarist and creator, posted a video about the issue. “It looks like something AI-generated,” he said. “It bothers me because it could erode the trust I have with my audience.”

Complaints about odd changes to Shorts surfaced on social media as early as June, but only after Beato and Shull spoke out did YouTube confirm the experiment.

YouTube admits to ‘limited test’

Rene Ritchie, YouTube’s creator liaison, acknowledged in a post on X that the company was running “a small experiment on select Shorts, using traditional machine learning to clarify, reduce noise and improve overall video clarity—similar to what modern smartphones do when shooting video.”

That explanation drew further criticism. Samuel Woolley, a disinformation expert at the University of Pittsburgh, said the company’s wording was misleading. “Machine learning is a subset of artificial intelligence,” he said. “This is AI.”

The controversy highlights a wider trend in which more of what people see online is pre-processed by AI before reaching them. Smartphone makers like Samsung and Google have long used AI to “enhance” images. Samsung previously admitted to using AI to sharpen moon photos, while Google’s Pixel “Best Take” feature stitches together facial expressions from multiple shots to create a single “perfect” group picture.

“What’s happening here is that a company is altering creators’ content and distributing it to the public without their consent,” Woolley said. Unlike Photoshop filters or social media effects, he warned, YouTube’s AI edits add another hidden layer between audiences and the media they consume—raising concerns about authenticity.

Creators respond

While Woolley warned of eroding public trust, Beato remained more optimistic. “YouTube is always working on new tools and experimenting,” he said. “They’re an industry leader, and I have nothing bad to say about them. YouTube changed my life.”

Still, critics say even minor retouching without disclosure sets a troubling precedent. YouTube is home not only to entertainment, but also to news, education, and informational content—areas where accuracy and authenticity matter.

The quiet rollout suggests a future in which AI may increasingly reshape digital media before users even press play.

联系我们 contact @ memedata.com