No password. No payment. No lock. No key.
In a California federal court, three YouTube creators — including the widely followed h3h3Productions — have brought Apple into a legal reckoning over a question that quietly underlies the entire generative AI era: does posting something publicly mean surrendering it to industrial use? Apple's response, filed this week, argues that what is freely visible to any person on earth is freely available to any machine, a position that, if upheld, would ratify a foundational assumption of how AI systems are built. The court's eventual ruling will not merely settle a dispute between creators and a corporation — it will help define the moral and legal geography of the internet itself.
- Three YouTube channels, including the millions-strong h3h3Productions, allege Apple systematically harvested their videos at scale to train AI — without asking, without paying, without acknowledgment.
- The lawsuit invokes the DMCA's anti-circumvention provisions, claiming Apple deliberately bypassed YouTube's technical safeguards, an act the creators call an unconscionable exploitation of a creative community.
- Apple fires back with a stark counter-logic: the videos were public, unpaywalled, and unpassworded — and public availability, it argues, is permission enough under both copyright law and YouTube's own terms of service.
- The same creators have filed parallel suits against Meta, Nvidia, ByteDance, and Snap, signaling a coordinated legal campaign to force a reckoning across the AI industry over how training data is sourced.
- Apple has asked the court to dismiss the case entirely, but the ruling — whichever way it lands — is expected to set precedent for creator rights in the generative AI economy.
This spring, three YouTube channels — h3h3Productions, MrShortGame Golf, and Golfholics — sued Apple in federal court in California, alleging the company had scraped millions of their videos without permission to train its artificial intelligence systems. The creators argued that Apple had deliberately bypassed YouTube's technical protections against unauthorized copying, violating the Digital Millennium Copyright Act, a 1998 law designed to prohibit exactly that kind of circumvention. Their complaint was pointed: Apple had built substantial value from their work while the creators received nothing in return.
The channels are not minor players — h3h3Productions commands millions of followers, and the golf channels have hundreds of thousands each. All three had filed similar suits against Meta, Nvidia, ByteDance, and Snap, suggesting a deliberate legal strategy aimed at the broader AI industry's data-sourcing practices.
Apple's response, filed this week, offered a fundamentally different reading of the law. Because the videos were posted publicly — no password, no paywall, no technical lock — Apple argued it had every right to access them, and that YouTube's own terms of service supported that position. More pointedly, Apple contended that the creators had failed to show YouTube's technical measures actually controlled access to the videos in the way the DMCA requires. The content was simply available to anyone online, and Apple argued that public availability cannot constitute circumvention. The company asked the court to dismiss the case.
At its core, the dispute forces a question the law has not yet cleanly answered: does making something publicly visible on the internet amount to consenting to its use as raw material for AI systems worth trillions of dollars? The creators say no. Apple says yes, at least under the law as written. The court's answer will carry consequences well beyond this case.
Three YouTube creators filed a lawsuit against Apple this spring, claiming the company had systematically scraped millions of their videos without permission to feed its artificial intelligence systems. The channels—h3h3Productions, run by Ethan and Hila Klein; MrShortGame Golf; and Golfholics—brought the case to federal court in California in April, alleging that Apple had deliberately circumvented YouTube's technical safeguards against downloading and copying video content. The complaint was blunt about what they saw as the violation: Apple had profited substantially from their work, they argued, while the creators received nothing.
The lawsuit invoked the Digital Millennium Copyright Act, a 1998 law that makes it illegal to bypass technological protections designed to prevent unauthorized copying. According to the creators' filing, Apple's actions amounted to "an unconscionable attack on the community of content creators whose content is used to fuel the multi-trillion-dollar generative AI industry without any compensation." h3h3Productions carries millions of followers; the golf channels have hundreds of thousands each. All three had filed similar lawsuits against Meta, Nvidia, ByteDance, and Snap, suggesting a coordinated legal strategy to challenge how major technology companies source training data for their AI models.
Apple's response, filed this week and reviewed by MacRumors, took a different view of the law entirely. The company argued that because the creators had posted their videos to YouTube as public content—no password required, no payment wall, no technical lock—Apple had every right to access them. YouTube's own terms of service, Apple contended, permitted such access. More fundamentally, Apple said the creators had failed to establish that YouTube employed technological measures that actually "control access" to the videos in the way the DMCA requires. The videos were simply there, available to anyone with an internet connection. That public availability, Apple argued, meant no circumvention had occurred.
"Plaintiffs allege that they posted audiovisual works to YouTube, and that any member of the public can see them there," Apple wrote in its filing. "No password. No payment. No lock. No key." The company acknowledged that YouTube does employ some technical measures to prevent unauthorized downloading, but said those measures were irrelevant because the videos themselves were not access-controlled. Under the law, Apple argued, the plaintiffs had simply failed to make their case. The company asked the court to dismiss the lawsuit entirely.
The dispute cuts to the heart of how artificial intelligence companies source training data and what obligations they owe to the people whose creative work fuels their systems. The creators' argument rests on the idea that posting something publicly does not amount to granting permission for industrial-scale copying and reuse. Apple's argument rests on the idea that public availability is permission enough, at least under current copyright law. The court will have to decide which interpretation holds. Whatever the outcome, the case is likely to shape how technology companies approach data collection for AI development going forward, and whether creators have any legal recourse when their work is used to train systems they did not consent to.
Citações Notáveis
An unconscionable attack on the community of content creators whose content is used to fuel the multi-trillion-dollar generative AI industry without any compensation— h3h3Productions and co-plaintiffs' complaint
Plaintiffs allege that they posted audiovisual works to YouTube, and that any member of the public can see them there. No password. No payment. No lock. No key.— Apple's court response
A Conversa do Hearth Outra perspectiva sobre a história
Why does Apple think posting a video publicly means they can scrape it for AI training?
Apple's argument is that if something is publicly accessible on the internet with no technical barrier—no login, no paywall—then accessing it doesn't violate copyright law. They're saying the DMCA only protects against circumventing locks that actually control access. A public YouTube video, in their view, has no lock to circumvent.
But the creators posted their videos for people to watch, not for Apple to copy millions of them at once for AI training.
Exactly. That's the creators' position. They're saying there's a difference between allowing public viewing and granting permission for industrial-scale scraping. YouTube has technical measures against downloading, but Apple apparently found ways around them. The creators see that as circumvention; Apple says it doesn't matter because the videos themselves are public.
What's at stake if Apple wins?
If Apple wins, it essentially means any publicly posted creative work—videos, articles, images—can be scraped and used to train AI systems without the creator's consent or compensation. The precedent would be enormous. Every tech company building AI would have a legal roadmap for data collection.
And if the creators win?
Then companies would need explicit permission or would face liability. It would slow down AI training and force companies to negotiate with creators or find other data sources. It would also establish that public posting doesn't equal blanket permission for reuse.
Why are these three channels suing multiple companies at once?
They're not alone in this fight. Filing against Meta, Nvidia, ByteDance, and Snap suggests they're trying to establish a pattern—that this is industry-wide practice, not isolated to Apple. It strengthens their argument that this is systematic exploitation.