Apple defends YouTube video scraping as lawful under DMCA 'public access' rules
Apple has filed a motion to dismiss a class-action lawsuit from YouTube creators who allege the company unlawfully scraped their videos to train AI models. Apple argues that the content was publicly accessible on YouTube without technical barriers, contending that such data collection does not violate the DMCA's anti-circumvention provisions.
Key Takeaways
- Apple contends that YouTube’s anti-scraping tools are usage restrictions, not the 'access controls' defined under DMCA Section 1201(a).
- The lawsuit, led by Ted Entertainment and h3h3Productions, alleges Apple used rotating IP addresses to bypass YouTube’s download protections.
- Plaintiffs cite an Apple research paper referencing the 'Panda-70M' dataset—derived from millions of YouTube clips—as evidence of the scraping.
- Apple seeks a full dismissal with prejudice, asserting that 'public means public' for content accessible without a digital lock or key.
Why It Matters
The immediate implication is a high-stakes test of whether platform-level anti-scraping tools carry legal weight under the DMCA. If Apple prevails, it effectively greenlights the mass ingestion of public video content for AI training, provided no 'hard' access barriers like paywalls are breached. For the streaming ecosystem, this creates a split: platforms may feel pressured to either gate content behind logins or accept that public libraries are 'freeware' for model developers. Watch for the judge's ruling on the distinction between 'access' and 'use,' which will set the technical standard for AI data harvesting.
Additional Context
The litigation against Apple is part of a coordinated legal push by creators against several major AI players. Per CNET (April 2026), the same plaintiff group—comprising Ted Entertainment, Matt Fisher, and Golfholics—filed parallel lawsuits against NVIDIA, Meta, ByteDance, Snap, and Amazon. These suits frequently allege that companies use virtual machines and rotating IP addresses to evade detection while downloading millions of URLs daily. In the NVIDIA case, internal documents reportedly showed employees discussing the massive aggregation of video data for generative modeling despite internal concerns regarding the practice's legality. While Apple and its peers fight these claims in court, other video platforms are taking a protective stance. Per PetaPixel (May 2024), Vimeo CEO Philip Moyer announced the platform would explicitly prohibit generative AI models from being trained on its hosted videos without 'explicit consent,' positioning the company as a creator-first alternative to the open-web scraping model. Vimeo also implemented security protocols specifically designed to detect and block automated data mining from AI model companies. YouTube itself has recently overhauled its policies to handle the surge in AI-related activity. According to Tech-Now.io (July 2025), YouTube launched a significant redesign of its Partner Program to demonetize 'AI slop'—low-quality, mass-produced content with minimal human input. While Google promotes its own video generation tools like Veo 3, the platform is simultaneously tightening restrictions on third-party automated re-uploads. In current copyright litigation, Google has argued that its terms of service grant it a broad, irrevocable license to create derivative works from uploaded content, a defense that closely mirrors Apple's current stance on the public nature of the service.
Read full article at hothardware.com
Get this in your inbox → Subscribe
Enjoy our coverage?
Add StreamingMeme as a preferred source on Google to see more of our streaming news at the top of your Search results.
Add as preferred source