The Verge - Artificial Intelligences 2024年09月06日
YouTube is making new tools to protect creators from AI copycats
index_new5.html
../../../zaker_core/zaker_tpl_static/wap/tpl_guoji1.html

 

YouTube正在开发新工具,旨在让创作者对使用生成式AI复制其声音或形象的内容有更多控制权,同时承诺打击非法行为并加强保护。

🎤YouTube推出合成歌唱识别技术,可让艺术家和创作者自动检测并管理利用生成式AI模拟其歌唱声音的内容,该工具位于现有Content ID版权识别系统内,计划明年进行试点测试。

👤另一项正在开发的工具可识别平台上创作者、演员、音乐家和运动员的面部深度伪造,系统仍在积极开发中,推出时间未明确。

🚫YouTube承诺打击任何人非法抓取平台内容以构建AI工具,采取阻止访问和投资检测系统等保护措施,强调AI应增强而非取代人类创造力。

🎁YouTube还在开发让创作者对第三方AI公司如何使用其平台内容有更多选择的方式,并将在今年晚些时候分享更多细节。

Illustration by Alex Castro / The Verge

YouTube is developing new tools that aim to give creators on the platform more control over content that copies their voice or likeness using generative AI. In its announcement post, YouTube said the new likeness management tech will help to safeguard its creators and partners while enabling them to “harness AI’s creative potential” by promoting responsible AI development.

The first tool, described as a “synthetic-singing identification technology,” will allow artists and creators to automatically detect and manage YouTube content that simulates their singing voices using generative AI. YouTube says the tool sits within its existing Content ID copyright identification system and that it’s planning to test it under a pilot program next year.

The announcement follows YouTube’s pledge last November to give music labels a way to take down AI clones of musicians. The rapid improvement and accessibility of generative AI music tools have sparked fears among artists regarding their use in plagiarism, copycatting, and copyright infringement. In an open letter earlier this year, over 200 artists, including Billie Eilish, Pearl Jam, and Katy Perry, described unauthorized AI-generated mimicry as an “assault on human creativity” and demanded greater responsibility around its development to protect the livelihoods of performers.

A separate tool is also in the works that can identify facial deepfakes of creators, actors, musicians, and athletes on the platform. The system is still in active development, and YouTube hasn’t indicated when it’s expected to roll out.

YouTube is also pledging to crack down on anyone scraping the platform to build AI tools. “We’ve been clear that accessing creator content in unauthorized ways violates our Terms of Service,” the platform said — which hasn’t prevented companies like OpenAI, Apple, Anthropic, Nvidia, Salesforce, and Runway AI from training their AI systems on thousands of scraped YouTube videos. The protections against this activity include blocking scrapers from accessing YouTube and investments in scraping detection systems.

“As AI evolves, we believe it should enhance human creativity, not replace it,” YouTube said in its announcement. “We’re committed to working with our partners to ensure future advancements amplify their voices, and we’ll continue to develop guardrails to address concerns and achieve our common goals.”

YouTube also says it’s developing ways to give creators more choices regarding how third-party AI companies are permitted to use their content on the platform and will share further details later this year.

Fish AI Reader

Fish AI Reader

AI辅助创作,多种专业模板,深度分析,高质量内容生成。从观点提取到深度思考,FishAI为您提供全方位的创作支持。新版本引入自定义参数,让您的创作更加个性化和精准。

FishAI

FishAI

鱼阅,AI 时代的下一个智能信息助手,助你摆脱信息焦虑

联系邮箱 441953276@qq.com

相关标签

YouTube 生成式AI 内容保护 AI发展
相关文章